I have a gold star here waiting for the first comment to tell me this video is too long! btw: This is just an hour of me talking. So I strongly recommend listening while you draw or something!
I studied 2D/3D at artschool 10 years ago. I always drew and painted and it seemed like taking those passions to the next level. However, I did like the drawing part far more than 3D. I used 3D to shortcut my concept artwork, for example for shading or general shape language for weapon or vehicle concepts. And it works well together, but it doesn't feel like drawing anymore. I later worked in multimedia, mostly in video editing. Last year I quit my job, because I was burnt out. It just didn't interest me. And it had made me lazy. I went to work and binged Netflix for ten years. I had completely stopped drawing. But last year I picked up a new Wacom, started drawing again and haven't stopped since. And my ambition is to be like you: I really WANT to learn anatomy and building poses, because I enjoy being able to draw my own comic. In my own style. What I have found is that it is more work than sitting in an office and doing a job. And part of making art is failure. It's not a task, like any other job I did. You can't be good at art without making mistakes. My truth is you make art by not giving up on yourself. Eventually it will become like a second language. Even though I am not there yet, I am building up to it and that motivates me to practice.
I work in the film and TV end and its important to understand that the people producing entertainment hate artists. They don't hate us personally but they hate the fact that they need us. This is what drives their fantasies about AI. Its the same reason the studios promote overseas production, motion capture, CG and now, finally it's AI. They've always fantasized about having slaves who do their bidding without making demands in exchange. This will never change. Art and business are distinctly different processes and its important for artists to learn aboutboth in order to protect ourselves from business folks whose interests have nothing to do with art.
This seems to be the intractable position. Perhaps the only hope is that artists can possibly create their own product and automate more of the business end rather than the creative process.
You seem to have similar views to Craig Mullins. He said in interview with Bobby Chiu, that if you ask bunch of artists to work with very simple basic tool ( like digital round brush or pencil) they will give you vastly different images that reflect their personality but if everybody uses complex object oriented tools it will all look similar. That's where the danger is.
People always ask me about my brushes and think it's something special but i literally haven't changed anything for years lol.. and do almost everything with a simple pencil like brush. Yet my art technique wise changes a lot, because a simple brush can do so many things. I am not a big fan of fancy brushes unless necessary..
@@yinyun2715 I don't think it's possible to post link in youtube comments easily. I tried and the post disappeared. You can search in youtube video called "Craig Mullins interview about the history of digital painting" and it was submitted by Rando Archive. The portion of the video I'm talking about is around 18 minutes where Bobby asks about technology.
@maboroshiiro yeah, those people are annoying i dont respond to them at all thats why alot of art is derivative because of brush beggars and the copying and imitation of others works asking for your process instead doing there own thing
When I first clicked "CREATE", I stopped and reflected.... I remain eternally grateful for every artist that has spent their energy, heart and soul to CREATE ART that has now been stolen. AI now reaches into the barrel like an ungrateful, ignorant child poised to create another pail, lifeless cluster of pixels pretending to be authentic art. The unfortunate reality is that this reflection will continue to be a rare experience by those who see art as a commodity rather than ART. THANKS TIM, Great video!
Neither Josh nor Tim knows the real right answer, that's the difficulty with it. Possibly ignoring it is dangerous to your career, maybe using it too much is what'll stunt you. I suppose everyone just wants to be told what they want to hear.
@@Balloonbot How the f*ck is it dangerous to the career? If I can draw it without AI involved thanks to my skills then I'm good. And using AI would be a waste of time that could have been spent on honing my skill
@@dailydoodle42 I take "ignoring it" as not paying attention and being uninformed of what AI is doing. Though if balloonbot means "ignoring it" as "not using it" then I absolutely agree with you. It's always going to be important to pay attention and be informed of GenAI's tech advances in order to understand what dangers it poses for artists and everyone else, doing so otherwise I think is irresponsible even if it's a dystopian black hole that sucks the joy out of life on a daily basis. But 'adapting' to GenAI by learning how to use it is just shooting yourself and your peers in the foot by validating its existence.
I used AI for a background in my comic to speed up my workflow. I already know how to draw backgrounds. I don't see the issue with using AI to speed things up. Some people bash on AI for the sake of bashing AI.
Wow people are really upset about AI. I spent 3 hours on my game art. I could have used ai to generate it in 2 seconds. Why did I go organic and draw everything by hand because I love it. Because my art 8s imperfect and that what makes it cool.
But generating in 2 seconds is how you get slop. What you don't get is you can spend more time on it to make it look good and different. You're just arguing against low effort, which nobody argues for.
@@LutraLovegood and even then it will lack original vision. It will never match the artists imagination because it has to draw from existing material and the random element really screws with the outcome
I genuinely considered the technology as being potentially useful, but at every turn I just realized how much faster I could do things myself. This conclusion has just made me more confident in being an artist, valuing the stuff I like more like stylized line art and construction to the point where I suddenly do feel like have a unique style with just graphite.
From someone who gets a peek behind the curtain... "AI" is a crock of shit in every way. I've seen an apt way of describing it: Burning down a forest to create the energy needed to generate an uncanny image of a tree. We already have AI, since before electronic computers, and that is corporations.
Tim you were my teacher at cgma academy back in 2012 i was such a noob at drawing but you teach me with patience and clarity, i am so thankfull to find you again in this platfprm, so happy to hear what you have to say as a real industry veteran. A big hug from Ecuador.
Really appreciate the full discussion here, we have differing views regarding Web 3 and use of AI at all and the like- I do appreciate that you’ve tested the tools to put the claims to the test in order to prove that they’re ultimately lacking - but regarding the actual benefits of having the skills internally, seeing students trying to circumvent the skill building, and coping with the wall in front of you, and AI being a way of getting there, it’s 10,000% my read on the situation as well. The needs that come from visual storytelling, posing, angles, hard and soft surface objects, push you to improve the inherent skills required. In the case of this particular video, some simple study of hair and hands would have gone far further than AI reference could. But when the majority of the work is typically heavily referenced, and there isn’t time in-between pinup illustrations spent on the vegetables of gesture and construction. It’s alluring to have the appearance of being an artist to others, but it’s far different to actually live the life and do the work that an artist does. Professionals have to figure these things out because the standards they have to meet are far higher, and specific. Spending 111 hours on a spot illustration or trading card would quickly lead to bankruptcy, and likely frustration from your Art Director. I know what it’s like to be making things FOR RUclips and an online audience, and I’m by no means the most skilled artist out there, but I’m also grateful that my career in the industry has provided plenty of hard knocks and reality checks that have fostered growth.
There is a basic point here that has to be said that many 'artists' out there already steal without any compunction. I make comics, and we had a variant submitted that was a direct trace of a famous artists work. If we published it we legit could have been in some kind of legal trouble were the owners of the art to press charges. "Trick" based art is so prevalent on social media that people think they can amass a series of tricks and become legitimate artists. And of course, non-artists have no clue. My writer is often completely fooled by artists who have NO foundation whatsoever and heavily rely on crutches even though he would consider himself to have an educated eye. I'm happy to see 'artists' who have an ends based approach without curiosity for the craft of art be atrophied from the field. On the practical side however, you make a point that in order to avoid being replaced artists should learn fundamentals. I 100% agree, but I will say also another practical thing is that the point you make about how it's "not the vision they had for the piece" doesn't matter to most people. People consume art in various methods, they don't consider it very much at all and will not care about whether something was made in accordance with art direction or not. For established artists who have a style and the technical capacity this won't matter. But there will be a hollowing out of the arts because there is a suppressed demand for low and medium tier artists. Why hire a bad artist to make something that isnt good when I can get good enough and very near professional with ai. As a result people who could have been great artists but needed a runway might not see their talents blossom. That's my current apprehension about the long term effect on the arts.
I'm pretty pro AI and use AI quite regularly is all of its forms, but I HEAVILY agree with your assessment of the effect AI will have on, not just art, but a lot of fields. AI is really good at doing the bottom barrel tasks. You want a quick render of your D&D character? done. Yo want a basic webpage HTML outline? Done. A lot of the tasks that entry level and even some mid level skill jobs did, AI can do. This is going to create a skill gap where employers will be expecting skilled workers to be able to come into the field at mid level or higher, and the vast amount of entry level skilled workers without a path to practice their craft to get up to the mid skill level. And soooo many employers lack the foresight to see that the short term gains of switching to AI will be offset in 10 years when the talent pool dries up. They are literally putting all of their eggs in the basket that AI will continue to develop at an exponential rate and be able to do the mid to high level jobs in 10 years. And if they are wrong, the entire infrastructure collapses as the high skill level jobs retire out and and there is no one left to fill the gaps at an affordable rate.
@@miclowgunman1987 It's an effect I don't see being talked about enough. Many want to make moral and legal arguments, but that entirely relies on your ability to enforce your morality or legal defense. Most artists are completely powerless, even industry leaders probably. So when artists complain about art theft i say...yes and? This is already a thing. People trace, or literally use the same images and most artists have no capacity to defend their works. I see this particularly because I definitely needed the runway. I needed to limply crawl before I could run. Someone had to take a risk on me, develop my skills and still make a profit. Who will be taking such risks in the future? If today it's 80% of creatives, but tomorrow its 70, then eventually down to 25% who only hire the best humans. Not a winning trend if you enjoy novel creation.
This reeks of infringement and not what it sounds to be--just another innovation--through the abuse of technology. The results cannot be extrapolated to other groups without the usury of infringement, to ignore everything we know is unreasonable. I see the loophole--just another innovation it isn't but covert infringement it is--we'd give way to new innovation but that's not what we're seeing here when it is a breach [edit: of] well defined boundaries commonplace in everyones hearts and minds [edit: even, or, how business is done. how people do the business [with one another.]]
It's a bit of a proletariat argument, but someone needs to build the houses for artists to live in, someone needs to make them lattes and bagels, someone has to build their plumbing. AI has ruined the idea of mundane jobs being replaced by coming after artists, but please let's also understand that while we'd all wish to see more genuine, productive (original), well developed artists, the notion that someone should be sponsoring them minimizes the labour of others. I'm not a fan of AI slop and the obvious ways AI is being used, but I still believe that many people who could not afford exposure to low quality entertainment, education, support and tutorials can do it now. Saying that it is low quality as an argument ignores the fact that many only have terrible quality available to them now.
using AI as a developing artist seems like a great way to avoid developing actual artistic skills, rendering abilities, the ability to observe and understand space, composition techniques, visual vocabulary, or any of the other things that allow a person to create images. most of the folks I hear talking up AI art seem to either be the sort that want images quickly without having to wait for them to be created (or you know, pay for them), don't really care how the art looks, or they literally can't tell the difference between well-rendered things and poorly rendered things. >:(
@@h.vallis6934 If you need AI to solve creative problems, you're not being creative. You come across as someone who has no understanding of their own creative process, or the value and joy to be found in difficult work. Art is not about the end result, it is about how you come to understand your own inner life through the tedious bits of the work that must be done to arrive at an end result. And no, writing prompts doesn't count...none of the output you get from generative imagery software reflects a decision you made, an experience you had, or a problem you had to question. You tune your prompt until you find something that is aesthetically pleasing, and that will always leave you with a Product that lacks authenticity. Just start drawing, dipshit.
How many times have you drawn a thing, and you absolutely KNOW that it is garbage, but everyone around you tells you that it's amazing. like 90% of AI art users want a quick picture of the thing they are thinking about. Pikachu riding a dog? done. a D&D character portrait. Done. These things don't need to be perfect and will likely be discarded shortly after. They are also very likely to have never paid for that image in the first place. Its quite literally giving people access to art(the commodity) that they want without all the red tape that comes with finding and commissioning(paying) an artist. They are both the creator and the consumer, and if the consumer doesn't care about the quality, why should the creator. That being said, all the things you have listed are absolutely required if you want to make GOOD art with AI. You more take the role of an editor, sifting through the many submissions the AI gives to try to find the proper lighting, coloring, and spatial proportions. The more advanced AI platform even allow you to directly choose these things through inputs of styles, 3d models, and other controlNet inputs. The better the artist is, the better they will be at controlling the AI to refine its outputs and in choosing a final product. Its akin to a camera: everyone can take a picture, but every picture isn't a masterpiece. Just because you can point and click, doesn't mean you don't need an intimate knowledge of perspective, lighting, colors, focus, and lenses, but technology can bridge that gap a good bit.
This is why people who aren't artists think generated images can be used to "speed up" making art by automatically doing the "boring stuff" with the push of a button. Like generating a bunch of starting ideas. People who refuse to learn how to create things never come to understand: the initial experiments are a necessary part of the process. You HAVE to make those yourself to warm up and begin subconscious thought processes that lead to the final creation. Even looking at photographic references is not the same - yes, those are existing starting points. But anyone who illustrates takes a photo reference and re-draws it themselves, to experiment with what they are seeing, pick and choose aspects to pull out of it. There could be an argument that the generated image is "just a convenient replacement for photographic references". Except photographs of reality don't distort it. That's the point of using photographs as a starting point before the artist begins abstraction. A photograph of a real lake is infinitely more valuable than an AI generated images of a lake, which is just a copy of a million distorted copies. And don't even get started on a AI generated image of a person which is mean to be "photo real". Starting with that as your reference for rendering the human image in art is a terrible idea. What about fantasy? Things that don't exist? Here's the thing. Every piece of art depicting a fictional thing is based on real things. When you paint a dragon, you reference real animals and real biology to understand what you are stitching together to create the fantastical creature. A photograph of a real crocodile is more valuable than an AI generated array of "dragons".
You nailed the main problem, AI is a way of cheaply commodifying art. The main issue for me is that in the long run, a lot of art forms will lose their value.
Sure, if your only metric for the value of art is how much money it can make. An art forager who can fool museum curators also calls the monetary value of art into question. Our collective visual culture is already largely defined by corporate work and people who want to cut corners. There's something deeper at the heart of all creative work that persists and is more important than the consumerist view of art.
are we just going to ignore the fact that Fievrr was a thing long before AI became decent? Art has always existed as both a commodity and cultural expression. Art as a commodity has been racing to the bottom since automation became the norm in factories and digital art made making art cheap.
@@plaguehaven but AI music is soulless. AI can only learn with really big sample datasets of a particular genre. At least for the foreseeable future, it won't be able to reproduce the soul of an artist
Though I'm just starting out and trying to build a professional career, I'm pretty much aligned with you on all points here. For the most part, I hate over-complicating my process, it sucks all the joy out of it, and though I rely on photo and 3D reference for very specific things (mainly historically accurate airplanes, which I draw a lot of, anything else I do mostly from imagination), I'm one of those people who shoot my own references (3D or physical models) because I find my creative process limited by trying to find the perfect reference, it drives me mad for the simple reason that it just never quite matches the vision. Very much enjoyed this take, one of the most grounded ones I've heard on this topic. Much appreciated.
It’s funny that you mentioned people sketching cars by making a 3D model and tracing it. Here I am, an aspiring automotive designer, having always strongly believed that 2D pencil sketches is where conceptual design should start and be refined. When I look at others conceptualizing designs in 3D, I feel as though I am falling behind the times, focusing on the wrong thing when it used to be the other way around. You’re the first person in a long time to say that 2D sketching should be the foundation and where you should develop your skills. This needs to be practiced by everyone who wants to be a designer or artist because that’s where you learn to create emotion in your lines, designs and art.
@@DuckieMcduck Not all artists are designers but designers are most certainly artists... it takes artistry and the result of the process is art. Art can be something as simple as moving a rock to be in an interesting place and telling somebody to look at what you have done.
the biggest thing is that in many jurisdictions the copyrightability of the product of AI is not clear, but in the US the copyright office has been clear that the product of AI work is public domain. If you're incorporating public domain content directly into your work through "AI tools" there will be a question of disclosure because how can you sell that public domain work without letting the client know they don't own parts of the image? How useful will your character design or spaceships be if they are public domain and everyone can use them? A lot of risk involved in AI - much more than in 3d.
it's equivalent to saying cowboy bebop isn't original because it uses bruce lee's movments for the mains character. I mean, you can totaly uses free and public domain art to creat new and original artworks. No art comes for nowere, even drawing for imaination, it is a blend of a lot of things you already seen in the real world, natural things or other people's artwork.
What you’re describing would be quite literally impossible to verify. Also, Marvel uses generative Ai. They don’t have any issues owning the films and TV they produce. The idea that Pixar and Transformers movies using Ai to do all their inbetweens will be unable to be copyrighted is just an absurdity, it’s a pipe dream that craftsmen cling to. The reality is, MOST media we see now and into the future have some measure of Ai integration. It’s not relevant to copyright. The few instances we’ve seen have been when people have EXPRESSLY and PURPOSELY tried to test the copyright system. No one need ever do that. It’s no one’s business how something is made. The reason that chimp photo isn’t copyrighted is SOLELY because the guy sold the image AS “a chimp takes its own photo.” Had he never mentioned that, it would have been able to be copyrighted like anything else.
@@WhatDoesEvilMeanyou totally misunderstand. Disney can use AI within a larger work & still have copyright, but that’s not the same as using generative AI to spit out a single illustration. If you do that, Disney still owns the character but what you’ve just created is public domain. That’s why when you freelance for them now you’ve got to sign promising you won’t use generative AI.
That's exactly what I've been asking this whole time, but seems that some tech corporation don't care to use generative AI into their workflow. Is STRESSING the amount of ads and video commercials that I've seen made entirely with generative AI, and I can't understand how are they supposed to represent their products or works if they CAN'T COPYRIGHT IT
This is the kind of conversation we need about ai right now. Really cool to hear your thoughts about it. The "don't let the tail wag the dog" is a great way of putting it. Real creative work should be guided by an artist's direction, and if you use ai incorrectly it definitely can derail your vision.
There is nothing that A.I. can add to my process. I love every step of creating my comic books from a blank page to a fully lettered page, I love every single step of the process. Sometimes I look at a finished page and I can't believe I did all of it.
I know that feeling I think this is why AI bros don’t understand what creativity is. They haven’t felt that dopamine rush you get when you’re truly happy with the work you’ve made.
Facts, I got into art because I eventually want to be that cool guy at a coffee shop drawing a person I see, AI can't help me get there and I feel there's a ceiling to it before everyone starts looking for real artists again with actual skills vs building AI prompts
The more AI becomes advanced, the more I realize it has no soul. Even when it produces intricate images of pure WOW, you can still tell it was generated by a machine, and not from a human imagination. I don't know why you can tell.
it's because AI does not make choices, or decide things, or even understand what it's doing. It has no understanding. It's parsing the aggregated data it's been trained on, sorting what in it's data matches the prompt given, weighs the percentages of how often prompted terms appear in images within it's data set, combines those things based on the percentage chance of those things appearing, and creates an image similar to those prompts, with some variance so that not every image is the same. some would say that this is similar to the creative process, but the key difference is that person who is creating actually has a plan of what they actually want to create in mind - AI has no plan, and will never be able to create it within the current AI framework.
@@bartelgrant When it was new and people were shocked and amazed by it. Most of those contest either don't let AI in now or have an AI specific category.
@@NightPhoenixPress not really man, pokemon trading cards had an illustration contest just a few months ago and tons of AI art made it within the top 100 submissions
I couldn't agree more...well said, Tim! It's still mind-boggling how many artists continue to take references from AI. We have beautifully crafted art books and incredible anatomy studies created by professional artists who have invested countless hours to provide valuable information for aspiring students. I still find myself revisiting great masters like Andrew Loomis, Norman Rockwell, and others, rather than paying for some AI site. Keep up the great content on your channel, as always!
In software, there's a phenomenon called Conway's law - which says, roughly, that the technology made by an organization is structured around the communication patterns within the organization: when teams collaborate on code, they write it around the organizational chart, even if not prompted to do so. Likewise, when you write code and then reuse the code later, you're "talking to your previous self", and therefore are limited by what they chose to do. And one of the problems raised by Conway's law is that you can't easily adapt the system you've built to a new context. The language you used previously informs the language you use later, which creates a cascade of effects on the software ecosystem as a whole. Sticking to a traditional tool like a pencil keeps you completely in charge of the language, while starting from within digital positions you downstream of the effects of Conway's law: there are thousands of decisions needed before you arrive at a painting app or a 3D model renderer. When these applications were new, they were considered extremely valuable, but when examined decades on, the benefits seem more dubious: most of the features feel like "lens flares and bevel filters" - a brief fad that quickly becomes a cliche. The identity of digital art has become more like programming in nature as the number of options to configure has increased with time. The AI tools being made now reflect a certain vision of using natural language as a command tool, with little acknowledgement of the unique properties of other languages like drawing or digital logic: those are subsumed into data. This quality has a parallel in how blockchain technologies are deployed towards a vision of tokenizing everything. Both are speculative: the language is almost certainly wrong, but has elements that could be built in a different direction. And I think that part of it is worth being excited about.
Very eloquently put. It’s true that using language prompts to generate fully finished illustrations seems doomed to fail as a result of the disconnect. Most artists will tell you that using words & making art are such different functions/ brain processes, that we don’t even like to do social media at the same time we’re painting. It reminds me of the bb king quote how he said he’d either sing or play but didn’t do both simultaneously 😂
@@scarletsletter4466 I feel it is important to note that the strongest AI tools in the market rely heavily on other aspects other than language. The language part is just the starting line. Then you supply 3d spatial poses for the characters, style sheets for the art style, pictures for lighting and angle references, and many more inputs. Then you have the ability to draw directly on top of the generated image and have it regenerate base off of the changes you have made. Purely language based AI generation is what all of the "Non artists" are doing. Its like saying all digital photography sucks because "look at all the bad photos people are taking with their phone cameras".
wouldnt the true foundation not be drawing with your finger on a wall with your own excrement? the comical part about the whole process argument is that it suggests there's only one type of pencil, on top of a single type of "foundation" paper able to hold any artwork imaginable, which is incorrect. you'll always go down the rabbit hole of "deciding downstream" if you wish to make "real art" in one way or another, and the most common thing to happen is being overwhelmed whether with digital or "traditional". most people dont know what some pens can do, and much also have no idea what their software can do, they find that one thing that works and stick to it. fortunately, you can't smell the digital.
Drawing is a non verbal activity- Artist's don't talk to their pencils- Drawing is the most direct and pure way to communicate a visual idea. I suppose some very clever tech bro could invent a technology that works by typing words in order to create pictures- but that would be stupid right? Because if words really could define pictures then we would not need the pictures and the blind could see just by having the world described to them- in words. Text to image is a genuinely funny idea because it fails at a very basic level, a failure that arises from a very human misunderstanding. As a species we are so entranced with language that we make the mistake of beleiving that words can actually capture reality. But you can't drink the word 'water'- and you can't really create visual art by using a bunch of words to describe it. Certainly you can approximate a visual idea by describing a picture using words but this is not making art, this is at best art direction, not making art. So the fate of the AI 'Artist' is to use words and a bunch of other tricks like 'image to image' in the hope that he can exercise some degree of control over his work- but in the end it's the machine that calls the shots, not the 'Artist.' Has anyone noticed that AI artists use the names of human artists in their prompts- or at least they used to until the AI developers stopped them doing it. But has anyone ever heard of an AI Artist using the name of another AI Artist in their prompts? No- me neither. I wonder why that is? ( Actually I know why that is- AI Art looks more like AI Art then it does the work of any particular user of AI.) The point is that 'Text to Image' is a really silly way to try to create visual Art, which is why AI Artists are never going to be able to deliver the granular control over their outputs that their clients will expect them to deliver. Any short term gains in 'productivity' from using AI art will in the longer term be offset by the basic intractability of trying to use words to define pictures.
As a writer, words can capture reality. A good writer can bring their writing alive in your imagination. "AI" art doesn't understand what words you are using as prompts, it doesn't understand anything. It only has key words which it scans, and spits out an image with, using a whole set of variables, loops, and conditionals that's programmed into it. Language, however, is the best conveyor of an imagination. There's a reason books have the broadest range of stories - words have no limitations.
@@magicbuns4868 I don't disagree with what you say here- words can create whole worlds of imagination. A book like Lord of the Rings for example creates a compelling mythological world that many people find deeply engaging. But what words cannot really do is express the subtle nuance of the visual. For example; If you wanted a portrait painter to paint your likeness would you send them a written description of your face? Or would you send a photograph? In this instance words alone would not be enough- no matter how detailed a description you wrote it would still be impossible to capture your likeness in the way a photograph does- which is why Passports have photographs. Another example might be a visit to the Louvre in Paris to see the Mona Lisa. If you arrived at the gallery to find that the Painting had been replaced by a detailed text description of the Artwork would you feel that in reading the text you would have seen the picture? Probably not. So you can see that the idea of using words to precisely define images is not really viable. Going back to Lord of the Rings- before the movies came out- eveyone who read the book had their own image of what Frodo or Gandalf looked like- your Gandalf, the one you pictured in your mind, probably looked different to mine. In fact each reader of the Lord of Rings had their own unique understanding of what the world of Middle Earth looked like. This is one of the great virtues of the written word- we can each have our own unique vision of the worlds of the story, precisely because the words allow us the freedom to imagine those worlds in our own way. But this lack of precision does make the idea of 'Text to Image' a bit silly, because if words cannot really define in detail what an image actually looks like then using words as a way to control an image making machine is inherently nonsensical. Those who claim that the images they get from their AI's are 'exactly' what they had in their mind are really engaging in a form of self deception in my view- they are in effect retrofitting their expectations to whatever the AI gives them- like getting a pair of socks for christmas then claiming that socks was exactly what you wanted all along! In this sense AI Art is a bit of a con- no one can really control what these machines produce with any real precision- the best you can do is get an approximation of your intent, filtered through the AI's training data and it's programming and further limited by the fact that no amount of words can really express in detail the image you are trying to create. AI Art works fine in situations where all you need is a crude approximation of your intent- but as soon as you want any real artisitc control they become more and more frustrating and the limits of a UI based primarily on words become ever more clear.
The ethical/moral/legal (and environmental) issues are obviously very important, but this message is critical too. Even for people who can manage to convince themselves that those issues aren't important, or that they're somehow enlightened and different, I don't understand why anyone would want to stunt their progress and shoot themselves in the foot with this kind of thing. Even if you never take it to the professional level, there's such joy to be found in exploring ideas and making choices, in learning how to learn and in learning about yourself and your style. In learning about your contemporaries as well as all who came before and in their process and choices. In filling pages with thumbnails! If people want to rob themselves of that growth, experience (and possibly community as well), that's their prerogative, but I do hope they understand the massive tradeoff they're making. Hopefully messages like yours will help inform.
I loved your take on this. I have struggled this past years not because of AI, but because I was in a bad state physically and emotionally. I was kind of burnout. So I had to stop, think about myself, about what I wanted, about the technology emerging, aka AI. And I arrived at pretty much the same conclusions. I always wanted to draw whatever I wanted, it was not about the end result, not even about me expressing myself, or my vision. But it was more about me being able to do it, If that makes sense. But many tools take away that feeling of "I did this, like yeah I completely did it" So I started working on my fundamentals again. And I understood that AI is designed to be very easy to use, even for normies. So in the case that the use of AI becomes normalized and accepted because all legal and ethical frameworks were properly stablished. I would be able to learn it really easily (I mean, the tech was like two months old and we already had AI "artist" and prompt "engineers", which I consider a bad use of language, but that is another story) and then I would have a pretty good competitive advantage. But if I can evade using AI and be competitive. I'll do it. I mean drawing is really fun! why would I be doing it of everything that I could if that wasn't the case? So using AI kills most of the joy.
Ai took low and mid jobs, which allows artists to do their craft at the same time paying their bills. Without that you can't get grow professionaly, while AI is getting better and easier to use. Let's face it they took our art without any consiquences and now they are selling blended pieces of our community labor. AI just made hard way of life of being an artist, even harder.
AI made my art better Because I realized how much my art looked like it ... It forced me to do the things that I noticed A lacked , that being : story telling , expressions , interactions , composition , im sure there's more. And yes, I use thumbnails now , great way to visualize the final piece , especially if your going for interesting color combinations
@@h.vallis6934 Yes it was. Digital art was/is (for the most part) simply a new media to use many of the same skills which are common to any visual art: composition, form, colour, etc. It differs from traditional media obviously, but there is still creation happening and you need to have an understanding of the fundamentals to create anything. Replace one artist with another, and different results are obtained even if the means are digital. With AI-prompts on the other hand you need absolutely no artistic knowledge or skill to "produce" an image on a screen. It is just a gimmick for non-artists to generate some images which allows them to feel like they get the results of being competent in the short term without actually being competent in anything. Theoretically speaking if two people type in the same words, they will obtain the same result. There is no artistic skill involved there, only a type of curation of computer-produced garbage (and as Tim explains in his video here, this is not even worthy of the status of "direction").
The problem for me is not relying on reference, but Josh over-relies on it. Like, he needed to pick a reference to draw an elf ear. Its not rocket science, its just a pointy ear.. also the hair on his illustration doesnt make sense, theres hair strands spawning out of nowhere, and that is the focus point of the image, very noticeable. doesnt look like he really took control of the 'tool'.
As an Artist myself, who started out with a Pencil and Paper in the 90s (now a 3D Artist in games) I completely agree with everything you've said here. Cheers! :-)
"The older I become, the more I realize that drawing is the most important of all the problems of picture making." Joaquin Sorolla - Spanish Impressionist.
It's a good point that if you learn how to just do it, then everything else is just a tool to get you there if you need it. Horrible ethics aside... In most cases I dislike the idea of using AI for reference because it's going to be a distorted version, and then it just runs into the copy of a copy of a copy problem of decreasing accuracy. I always felt like the most promising future generated imagery had was basically custom tailored entertainment. It's like a toy, not the same as seeing an artist's work which is a form of human communication. That's where I see a lot of conflict is in confusing those two things.
Wow you hit the nail on the head. The main thing I saw from the video from the artist you’re referring to is there was a big lack of having a simpler process along with not understanding that their issues were all foundational. It reminds me of something Will Weston said, that the problem a lot of artists have is that their methods are too cumbersome, and because many of those artists don’t bother learning how to simplify their methods and apply the fundamentals in a more simplified way (like a 20th century illustrator would approach it) they tend to compromise on their overall vision.
I tried creating art with AI when this craze was just starting about 2 years ago. Generated a bunch of sci-fi robot images then painted over them heavily. It ended up looking like something I would've painted myself. But the entire process felt so boring, there was nothing rewarding about creating that piece of art. Proving to me that sketching a bunch of ideas and working on them slowly is the best part of the creative process. "Maybe people can use it to create thumbnails or something" I thought, and that would be it. Never expected people writing full names of artists as prompts and generate uninspired versions of their beautiful artwork, basically stealing their years of experience.
This whole "drawing from imagination" certainly is a good exercise, but it's not what everyone strives for. You have to draw from reference in order to eventually draw from imagination. But when I say "draw from reference", I don't just mean draw what you see, but rather, understand what you see. Try to challenge yourself, like, what would the pose look like in a different perspective etc. However, I have to note here that the more stylized the artstyle is, the more simplified the forms are and thus the easier it is to draw things in perspective. Meanwhile others want a semi-realistic artstyle, which requires more knowledge about the human anatomy etc. It can take years and years of drawing in order to get to that point and references can make that easier for you. I still take photos of my hands if I'm not sure whether the hand looks right. I still look up references of arms and legs in specific positions cause I'm not sure if I forshortened them correctly. But what I agree on with you, is that unless they wanna draw realistic or copy an actuall reference, people should stop wasting their time finding a person that does the exact pose they wanna draw. It's not helpful, you will forever rely on references if you don't dare to try and challenge your imagination every now and again. At least in my opinion, but maybe others have good reasons why they continuously draw from existing pose references etc.
I agree. I think referencing is so important for artists to utilize properly. We really ought to move past shaming artists from using other images and photos as guides to better enhance their artwork. Drawing from imagination is largely just referencing from memory. The more reference you use overtime the more your visual library grows, allowing you to draw from imagination more reliably.
tbh, this video was something I needed. I'm someone who gets stuck in the same mindset and continuously worry about it. Same thing happened when I heard about Ai and artist. As an art student myself, I can't see myself in a different career path, so I started to panic. But somehow this video help me calm down from that mindset. Thank you
This was extremely relatable to me. Currently, I am an incoming junior into an entertainment design program and in the last year or so I have completely pivoted from illustration to 3d modeling. I can't deny that the programs limitations change what type of art I am creating. This video definitely helped me see what is truly valuable when creating art and I am glad im not the only one getting stuck on this.
Thanks for this really enjoyable video Tim, I really appreciate the balanced view of AI as it currently stands. I'm 7 months into my first design job inside a small company where it is obvious there is a fascination with AI from a business stand point. As you have said, if the AI could do my job - it would and I wouldn't be working but that still doesn't entirely calm my nerves about my employment in the future. I feel like I will need to try and expand my artistic incomes just so i dont have all my eggs in one basket
Saying that using a reference is the same as using AI, is like saying that using a recipe for a dish is the same as asking for someone to make it for you.
Not at all. Using a recipe is like following a tutorial. Using AI and references is like... watching somebody cook next to you, doing the same and saying you came up with the recipe lol
If you copy a reference exactly and say you came up with the image, that's just wrong mentality. I don't think a lot of people do that. References are used best for learning specifics, be it painting styles, lighting, colors, etc. Using images of a sunset to learn how lighting in sunsets work, that's following a recipe for sunsets, albeit in a more abstract manner.
Great video! I agree - I'm just tired of even hearing about AI art. It's not going to stop me from doing my own art, so I don't really care about it. I just wish it was easier to filter it out online, especially when searching for reference images. I don't want to be accidentally using AI art images as drawing references without realising it.
10:45 so I am a game developer and I do think LLMs are pretty great at bug fixing or writing code. The thing is, it isnt sustainable to make an entire game on its own. Its sort of like, you have something that can design machine parts, but it cant design machines exactly. So you can get small components basically for free, but you then have to understand how those components work in order to make the actual machine. And heck even to make the components to the right specifications, you need to know what those are. You are best off just doing as much as you can on your own, then trying an AI when you hit road blocks, if the AI cant fix your issue, ask someone on a discord. Its basically a replacement for stack overflow more than anything. Really I have noticed the same thing in programing as you have with AI art. I can waste a lot of freaking time if I try to use it for the wrong thing. I'd rather just use it for simple syntax or catching a programming error, and actually program everything myself.
The Director comments. I have heard successful Directors say, some days, some shoots, some scenes - they have no idea how it's going to work out. Sometimes it comes together in post, when they start to assemble. In fact, I've heard - from a storytelling point of view - entire movies have been rescued during the editing process.
That’s a way of humble bragging or saying they direct by instincts. They’re still a human who is making decisions with their brain. Even if we take those comments literally to mean they truly didn’t consciously “know” why they made decisions the way they did… there had to be some reason. Even if they were using a random number generator to come up with ideas, that’d still be a method
@@scarletsletter4466 Yeah, I think sometimes the creative process finds itself, in a way. Sometimes. I'm sure there are days when a strong vision from the start is followed through, right to the end, and other times it's more like play, where you're kinda just figuring it out as you go, following your instincts, fixing it in post, and experiencing the magic of 'happy accidents'. It can happen - and I've heard, directors, song writers, et cetera - talk about it. I've experienced it. I even heard a video game developer who worked on Half-life 2 describe the game coming together in a similar way. That was surprising to hear, but given how long Valve can take on a project, maybe that partly explains why. I'm just saying I don't know if I completely agree with Tim's take on the Director thing.
As a 3D artist, I use enhancing AI to upscale renders, upframe animations, or denoise images. That all leads to less render time needed to achieve my end result. When it comes to generative AI, I am basicly agaisnt it in all its forms. I mean, if somebody uses generative fill in photoshop like an advanced copy stamp, only generating meaningless background, it's lazy but if the hero object in the image is crafted without AI, I wouldn´t even notice the difference. ANd maybe it´s not even lazy in an age where we all need to be fast in execution. But since generative AI is used mainly to completely circumvent the progress of coming up with a visual concept and then flesh out the detals in painstaking work, I think it is just killing the need for skills. Funny, since the emergence of generative AI I have grabbed my pencil more often than in the last decade. I was focussing of 3D modeling stuff but that stable diffusion stuff really reminded me of the fact that I have neglected my drawing skills even before all that Ai stuff. I took it as a waking call for me.
I really needed this video, Thank You! I feel like I need to pay you a therapy session bill for this! 😅 I'm constantly anxious about AI as a freelance illustrator, and this video has helped me so much Tim, Thank you!💛💛🙏
There are entire worlds trapped inside a pencil, pen or brush. It's up to us to let them out. AI is not an achievement. It's buying food from a restaurant, putting it on your own plates, and telling your guests you made it yourself.
I've always thought AI would only help artists stop improving and you kinda touched on that. I don't think it's harsh, it's just the truth, you stop trying to improve your weakest spot, you won't get magically better. I never understood the claim that people who don't use AI will be left behind, either. Like we can't all learn to type a couple words in a couple minutes of messing with the stuff? People felt so pressured into it when it started to pick up attention, for fear of losing work, but it's made to be easy to use.
Just discovered this channel and decided checked out your online page. I love to draw and recently started exploring 2d animation and doing some fantasy writing. That being said, I’ve never really gotten past the hobbyist phase of my artistic journey as I struggle a lot with anxiety when it comes to my art. The moment you mentioned Asterix, Obelix, old video game instruction manuals, and Princess Mononoke on your ‘About’ page, I knew I was in the right place to get some guidance with my craft as those were all discoveries that really influenced my love for art through my high school and college years. Watching your free Mini Workshop now, and I’ve just subscribed! 🔥👍🏽🔥
I love long talking videos and your voice is very nice to listen to - also I am an artist and i enjoyed the vid! I'm not really interested in AI, but your video did solidify my own feelings and helped me clarify the way I communicate with others about art and AI in general. I'm a big proponent of using the simplest tools as creatively as I can :) Thank you!
Thanks Tim, I really needed to hear this today. I keep going from a doomsday scenario losing my job, to feeling like things will be okay. It's taking up way more brain capacity than I'd like, but I can't seem to push the negative thoughts away. I guess all I can do is carry on doing what I've been doing and hope clients still want to work with me over AI.
"Are we being luddites?" interestingly, the original luddites was a 19th century textile worker movement. these workers saw their bosses using profits generated by their work being used to buy automation machines that was going to decrease their wages and cut jobs, inxtead of creating the same level of output with less work for the same pay. If someone calls you a luddite, they're actually calling you pro-worker, someone who advocates for the principle that new technology should benefit all of us, not just the owners. This relates to AI in that very sense. If it was a tool that works, and it could help peoples workflows, instead of it being used to "save" on labor costs and cut jobs, it could be used to let the same amount of workers create the same amount of productivity with less work and the same pay, or increase productivity for increased pay. Capital's drive to generate profit regardless of the cost is what creates this tension between workers and new technology, something our predecessors the luddites understood.
Yeah... I'm aware of all that. It's an interesting story. I think the ship has sailed on the usage of the term though :) the winners write the history books. It's not a good look no matter how you 'well actually' the verbiage. We are best off learning from history to have another crack at standing up for pro-workers rights. It didn't end well for the Luddites...
People don't realize the release of Endorphins while making art with your own bare hands and at the same time building a skill that will most likely be gone in few years so that Art once again will be done by the elites and sold in closed off auctions.
You underestimate the endorphins people feel while being able to physically manifest the thing they were just thinking about into physical form just by typing its description into a box. There is a reason millions of people are subscribing to Midjourney, and millions more have downloaded Stable diffusion. And AI doesnt take away anything. Just as you still find people drawing and painting, even though digital art is easier and cheaper. The only thing that will shift is art as a commodity, not art as a form of expression.
@@miclowgunman1987 I’ve seen a lot stories of former ai bros who admit they stopped to enjoy “making” ai “art” even though they were getting results they wanted. It’s just not fulfilling in a long term. In my opinion, when you generate a picture, even if it’s everything you wanted it to be, it is more like randomly finding preexisting image on the internet, because you didn’t do anything except typing some words lol.
Great video Tim! Fully agree with everything you said about ai. It is just way too hard to get anything useful out of it. Having solid foundations and do everything yourself is the way to go. Using 3d on the other hand is super usefeul and changed the whole industry drastically as you said. The cool thing is that you can design and iterate pretty easily in 3d these days, so it is just a tool but not a crutch. You still need to be purposeful in how you use that tool. Ai is just too random to be useful
Thank you for being the voice of reason here!! 🙏 With the current takes of some bigger artists, I felt like even artists are forgetting how these genAI programs use stolen artwork 🤐 For me, the lack of usefulness to artists is secondary but seeing you speak out against AI fills me with hope regardless and I'm super thankful to have found your video! I hope even more people get to see this 🙏🥳✨
As someone who has contracted artists in the past, I choose people who have a consistent personal style that I like. Nowadays, when I see a portfolio from someone with different styles (and especially varied quality), I tend to assume the rely too much on AI art (or that the crappy drawings were them without AI, and that the suspiciously highly rendered images are AI doing most of the heavy lifting). That's a turn off for me. The pinnacle of working with an artist for me is chatting with them about the basics that make a character or scene identifiable and the impact I want the image to have... but let them create the scene that best show cases their style which I already trust. I don't have to be sold on the benefits of working with an artist over AI. I just have trust their art style and by sold on the quality of their communication/collaboration.
I love your style! Somehow someway I have never heard of you and you are so awesome and so is your channel! How have I never heard of you?? I need to buy your sketchbook asap!
The Silicon Vally hype culture.... hit it right on the head. People spew LOTS of nonsense especially in the film business right now because there's a lot of transition and things feel kind of dark right now.
I had a lot of fun messing around with Midjourney in the beginning before realizing I'd rather get back to drawing and that very few use cases of importance to me would be better served by using AI. I'm not ideological about it though, and I think there people making cool work with AI, usually by screwing it up in some way to create strange new images.
loved the video! im so glad you talked about the importance of learning the foundational skills and how vital it is and allways will be for a creative person, its what i told myself when i started to worry about AI emerging in art, and its so funny if you actually play around with A.I as an artist who understands how to construct an image from scratch using fundamentals how frustrating and boring it is to use and you end up thinking it would have been way quicker just to sketch it out. i myself use alot of 3d in my professional work when it comes to ships and vehicles but before i learnt 3d i knew how to illustrate and design a vehicle in perspective by hand also so im not 100% reliant on 3d software, just using it for speed when it comes to clients demands, i still sketch and paint freehand for practice/fun and jobs where i know i have enough time to do it all by hand.Again i think its so important to be grounded in the fundamentals and everything you said about that fact and parts about being a true director with a solid idea was spot on! glad i found your channel
I click this video expecting another "is better if you don't bc AI will take my money away" but im glad to find a real artist just talking about how is boring and lacks quality. As a 3D artist (that has drawn a lot too, which is harder if you ask me lol)I use it to TRY and get an idea out of my brain with it, but i fails more than succeeds most of the time, also for human reference is absolute trash, unless you want to draw the same girl with questionable anatomy 1000 times lol I have been doing this for almost a decade and some, and honestly AI is a fun curiosity, is kinda fun to get results from text but thats it, if you have a visual style you mostly will end up doing it yourself bc AI just makes it feel icky idk what it is but is very cliche all the time and since you didnt make it you cant iterate anything from it. Not taking into account it takes A LOT of tries to get barely close to what you want, so in speed I lack at least. I can see it making it faster at some point but if you have a quick mind with a good visual library drawings win 10/10, maybe some cheap cheap businesses can use it to sell pens or something but on the game industry and real good art like 2D animated movies, I dont see it making it too far, taking also in consideration some ethical issues, for now I used it to make fun AIs of my dog and some profile pics to avoid data scraping of my face lol, but I will try to make some content soon so my face will be all over the internet so not even that now lol. Anyway good video, you seem like a very passionate artist that enjoys doing it, which imho is the most important part why would you want to skip the fun part of life
If I had to make a video/talk about AI it would be the very same as what you've just done here. I am a professional illustrator, over 20 years of experience in various organizations and now, like you, in web3 and blockchain, because, well, we need to make a living. On the side I've always worked on my personal artwork and more interested in the history of fine art than gaming. Great passion for comics in my early years, though. My organization is pushing for AI more and more to the point it frustrates me and my colleagues. We see what they are trying to do. We use Midjourney at times to come up with references or sometimes even use its generations in the final, with due modifications. You can imagine how that makes me feel. First of all, AI never makes what I have in my head. It can be useful in industries with no artistic culture and seeking very standard aesthetics (most of them) and imitations of other things they see. In that case AI makes exactly that and it can be a shortcut. But when I work on my own pieces I am in a totally different mental zone. I came to feel when my drawing and sketches come alive and resonate with what I want to say. That "aliveness" is something that only those who dedicated years to a craft can understand, certainly not marketing managers or similar characters. But when an artwork goes through such a process then it does transcend and speak more. AI removes the process. That's why it can't be considered an artistic tool for illustrators and designers. By using AI you stop learning what art is, how to make it, how to feel it. AI can be ok to substitute non-artistic processes, like perhaps some tedious manual tasks of 3d production (standard rigging comes to mind perhaps), but not much else. My solution to all this is: if you are an artist, keep learning the fundamentals but also dive deep into where your heart is. Before using a reference, try to draw that thing as best as you can, until you bleed on the paper and only THEN find a reference and compare the two. Only that way you can internalise the process and strengthen your drawing skills. Don't fear mistakes, because mistakes can be inspiration. Imagine you live before the internet era and work as if you didn't have google nor many books to refer to. Use your memory and observation skills to create a mental library of anything is an interest to you. Let music, books, stories, people...anything really...inspire you, not just artists. Then your work will flourish, will have more and more soul, become more unique and no AI will ever be able to beat that nor will ever be a seductive alternative. AI has been created by people who don't know art. Art is a process. Artwork is its outcome. Live artistically.
I've done some art using illustrator and photoshop, and I drew a lot when I was younger. I'll admit my drawing skills have languished over the years, but I still 100% need to start an idea by doing sketches. Even if they arent something that I would feel would sell a client on my work, but rather just work for me as part of my conceptualization process to help me bring the fuzzy concept in my head more into focus and then have those sketches as a reference to draw from while using illustrator or photoshop to create a polished, high quality, professional, and fully realized result of my initial fuzzy/out of focus idea that I first had in my head. As soon as I can afford a GPU I plan on starting to learn blender, and I still plan on using pencil and paper to help me work out things like form and layout as part of my initial process before any 3D modeling takes place.
Hi Tim, I just found your channel and I appreciate you sharing your perspective! I've been a character concept artist in VFX for almost a decade so I thought of chipping in to your thoughts. I still think AI gen is at least a useful brainstorming tool even at this early stages. Yes, making 5 drawings by hand does build that muscle and is authentically human, but in that time an 'AI artist' can provide the client with 500+ images to choose from. Is the sheer quantity and speed of production that wins over authenticity and quality. (although image gen model are getting so good and realistic by now that only something like the top 5% of artists in the industry could compete in the value they can provide over it) I do agree it's worrying to think that humans will progressively lose their artistic craft and way of direct expression through it if industries will rather opt for the cheapest and fastest solution which currently looks to be AI. I also think that if the market goes in that direction, without intentional appreciation towards human made art, we'll all be at a loss from a human originality point of view. From a different angle though, having generative models which 'digested all of human made art' that is now in the hands of everyone sounds like a powerful progressive commodity which can indeed allow everyone to design, brainstorm ideas visually and create a world for ourselves that may overall be better than for humans to rely solely on the 10-20% of the population who dedicated their lives to mastering the craft and acquiring the skills needed to realise people's visions. I see generative AI in general as an immensely impactful technological innovation which for better or worse is causing a global multi-industry shift which can't really be stopped, but can be veered responsibly based on how we collectively carry ourselves.
I appreciate that you're looking at this from an artistic perspective rather than ethical, I've been feeling like I'm going crazy because people keep using generative images to inform their art and saying it looks good. It always looks bad. At best it's boring. I've tried and never been able to get anything that looks good. Ergo Josh's hair reference looks awful. It's just a big mess and it's reflected in the final piece. The parts he did himself are nicely rendered but other than that I don't think the piece is that interesting overall. And he took literally over a hundred hours to make that. If he had just looked at real hair and practiced for those hundred hours maybe he'd be able to draw hair himself.
I'm also a veteran in the comic industry (almost 20 years), but my dream has always been to create for the French comics/bande dessinée market. How did you reach them? I have an agent, but he doesn't have contacts in the French Belgian industry. Any tip? :)
love this! I start with thumbs and concepting the characters (usually for a series of illustrations for children's books). My work is easy when I am working with simple angles, but I do tend to use interesting angles for which there are rare photorefs. So I will digitally sculpt and sometimes use 3d environments to speed things up. I'll use AI to generate photorefs that I would normally have trawled to know what a thing will look like. I am about to get a 3d printer so that I can print out my references as I am always fighting my way to work as little digitally as possible. Sometimes its just about being fast and getting the concepts cleared. I don't render colours though - everything is essentially models to work from, be it 3d or I sometimes make models. James Gurney I think is the master of making models to work from IMHO. Great work as usual!
I respect your commitment to traditional art and fundamentals, they are important, and I love your art as it is inspiring. But I’m using AI generator inside Krita to fill clean canvas with something to start with. I don’t think that AI will be able to replace artists. It could draw generic assets, but not composition, and not finished compositions, that’s an artist job to guide it. People with vision would stand up, not the tools. But for now the level of control over AI is still very limited, and that’s the problem which creates divide between artists and tech
I wonder if the quality of art has dropped over the last century is because people aren't willing to put in the work? Is that why some just scoff and say, "Anything and everything can be art!" But if that's true, then nothing is art. Is this how we go from carved marble statues that look alive to making some large abstract statue out of clay? Has art fallen so far that people say the "meaning" of art is the only important thing, and create an "art piece" of scattered trash that resembles a wild party that never got cleaned up? Art has NO shortcuts! We mustn't get lazy, but I fear social media has affected art by forcing artists to churn out art faster than they should; "You can't rush art!" that old guy from Toy Story 2 said, and perhaps artists should heed that advice.
hey tim I hope you see this comment, I would really like to make art my career much like you do. there are dozens of videos about different jobs in the art field but when I look I just can’t seem to find any actual work. it you be awesome if you could share how you find work or how you found work in the past. love the content keep up the great work.❤
literally one hour of simply saying "have you tried picking up the pencil and drawing?" Completly agree wth everything though i'd say I'm even more harsh about AI as a useful tool. I don't ever see a computer replacing the human heart.
New to art as well 3d art. This video was really helpful. Thank you for making this. Almost all tutorials recommend sort out references as the first step. Using pure ref or something as the first step. And whenever I try to do that, i lose all interest in that project. Trying to use mid journey to create reference images kills the creative satisfaction completely. I always prefer using references or ai only when I am stuck with visualising something midway through my artwork, like unable to exactly imagine how does the foam around the rocks in a river looks, which is just a part of my painting, or how to get a the right lighting for a certain kind of look for a blender render. But starting with ai or a reference and trying to improve never works for me. I have been pushing myself to try to learn the reference first thing. And it was making me hate the process. Thank you for this.
IMHO most people don't care nor pursue the artistic side of an Artwork, be it digital art, architecture or concept art etc. Art is for them just stunning impactful visuals, and that is just AI-art is. A bit similar like fast food industry of art, only that Art isn't that essential like food for most.
I always found delightful to search for references among photographs. It's a way to broad your artistic view and improve your artistic taste, also learn to appreciate other people works. Like you, Tim, I have tried Midjourney on the past (when working as a graphic designer) and found it extremely boring. I prefer to use stock images and primp them.
I've had this exact problem of "letting the tail wag the dog," but with 3d instead of AI. I consider myself a 3d artist/animator, but I've recently realized that my 3d art is just... bad lol. And of course it is! I've skipped the fundamentals! So I recently decided to finally learn how to draw in earnest, and boy do I wish I had done so sooner. I'm having the time of my life just learning the basics. I'm looking forward to returning to 3d *after* I master the foundations and develop a unique voice so I can see how the tech will enhance my work. It's become clear to me that drawing is really the best and only way to learn and master the fundamentals of art.
I think people have an intuition that master the craft of art over 15 years of hard work is worth it even in the advent of a miracle box that does the work for you and it's not just the sunken cost fallacy. They are just waiting for someone to precisely articulate the reason as for why. I know that I am. I have not heard anything yet, that convinces me entirely. It is something deeper and philosophical.
also the situation is a bit more multifaceted than just the consumers nor caring, while it is true that most consumers dont care about how automated a particular industry is unless it affects them directly, the fact of the matter is that generative AI isnt just a threat to just visual artist but any job that requires expression through a visual medium, that means models, photographers, 2d artists, 3d artists, news broadcasters, actors, content creators etc, and not to mention the copyright cesspool that AI has caused with all the impending lawsuits. The fact of the matter is with all the industries that it threatens not just individual jobs, people will start rejecting this type of content in mass since its a tool at its core designed to undercut people. Eventually it will also cannibalize itself because if AI can be copyrighted while simultaneously feeding off people's work then eventually nothing will be safe within copyright law, since anyone can use your "ai work" and use it no one's work will truly be safe under copyright law. Also the huge potential for political misinformation with deepfakes and how it might affect evidence in law. SO huge rant aside, the consumers dont care for now but they will when it's a threat to almost every job. Sadly sometimes things need to get much worse before it gets better
I think directed image generation has its place in a support role. Image generation is really clunky when it comes to creative control, but for in-betweens where you mostly redraw the scene, for creating non-descript, unimportant details it could very well speed up a project.
I'm a concept artist in between jobs right now (and I realize I might not see the next job). I generated a knight that looked kinda cool in midjourney, fed that in Hailuoai that generated a 360 turntable out of it, fed that into a photogrammetry program, and it actually created a 3D shape that I can use however I want. Did the same for environments. I think we can look at concept art in two parts that's essential of what we do for a company. We help the creative process and as a by product, create things that other departments can work off of. It takes a lot of skill to generate 2D designs, and to translate them into 3D. Concept artists were used to solve, to varying degrees, for both issues. However I just demonstrated we can skip the whole process of a concept artist entirely and head straight to cleaning an end product. Do I want to do this in the future? No. I think i'm checking out and become a night guard, draw/create in my free time and will definitely not share art online.
That sucks, sorry to hear about your job. I totally get where you are coming from. This is not career advice... but I think the more you can focus on the actual design aspects of what a concept artist does the more you are going to secure your future as a professional artist. Many companies and projects have essentially been cookie cutter copies of each other for the last decade or two as games have become a huge industry. This derivative ideation is no longer valuable as AI can do a good job of regurgitating concept art sludge. But the technology is fundamentally built on the logic of serving up what it thinks should go next based on it's dataset and neural net. This means that the value of this type of work has gone to zero. But there will be more and more need for actual design and innovation at big companies. I think any large corporation leaning into AI for ideation is going to find itself easy competition. The biggest asset a large ip based company is going to have in the future will be it's ability to come up with actual new ideas and get them to the marketplace before they are copied by cheaper AI competitors. I think it's best to always see yourself as an artist separate from the jobs or career you do. The career will change... our role will change. but your artistic self and the stuff you are interested in should be something you continue to develop and build on. The only advantage of getting a job in a creative industry is you get more practice to build your skills.
Thank you for sharing your thoughts about this. Artists bring originality where artificial intelligence generated images are recycling what is already existing.
I have a gold star here waiting for the first comment to tell me this video is too long!
btw: This is just an hour of me talking. So I strongly recommend listening while you draw or something!
Idk, I am enjoying long videos like this. Nice to listen while drawing.
I'm listening while CAD-ing, but similar parts of the brain employed
I like longer videos so no complaints from me
I like the conversation, but I was kind of hoping to see you draw. LOL.
I studied 2D/3D at artschool 10 years ago. I always drew and painted and it seemed like taking those passions to the next level. However, I did like the drawing part far more than 3D. I used 3D to shortcut my concept artwork, for example for shading or general shape language for weapon or vehicle concepts. And it works well together, but it doesn't feel like drawing anymore.
I later worked in multimedia, mostly in video editing. Last year I quit my job, because I was burnt out. It just didn't interest me. And it had made me lazy. I went to work and binged Netflix for ten years. I had completely stopped drawing. But last year I picked up a new Wacom, started drawing again and haven't stopped since. And my ambition is to be like you: I really WANT to learn anatomy and building poses, because I enjoy being able to draw my own comic. In my own style. What I have found is that it is more work than sitting in an office and doing a job. And part of making art is failure. It's not a task, like any other job I did. You can't be good at art without making mistakes. My truth is you make art by not giving up on yourself. Eventually it will become like a second language. Even though I am not there yet, I am building up to it and that motivates me to practice.
I work in the film and TV end and its important to understand that the people producing entertainment hate artists. They don't hate us personally but they hate the fact that they need us. This is what drives their fantasies about AI. Its the same reason the studios promote overseas production, motion capture, CG and now, finally it's AI. They've always fantasized about having slaves who do their bidding without making demands in exchange. This will never change. Art and business are distinctly different processes and its important for artists to learn aboutboth in order to protect ourselves from business folks whose interests have nothing to do with art.
bingo
Exactly!
This seems to be the intractable position. Perhaps the only hope is that artists can possibly create their own product and automate more of the business end rather than the creative process.
Yea & on conference calls you can always hear it in how they refer to us as “creatives” like it’s a slur 😂
Very well said and probably one of the most important things artists must to always remember.
You seem to have similar views to Craig Mullins. He said in interview with Bobby Chiu, that if you ask bunch of artists to work with very simple basic tool ( like digital round brush or pencil) they will give you vastly different images that reflect their personality but if everybody uses complex object oriented tools it will all look similar. That's where the danger is.
Mullins was almost prophetic
People always ask me about my brushes and think it's something special but i literally haven't changed anything for years lol.. and do almost everything with a simple pencil like brush. Yet my art technique wise changes a lot, because a simple brush can do so many things. I am not a big fan of fancy brushes unless necessary..
Link for the interview?
@@yinyun2715 I don't think it's possible to post link in youtube comments easily. I tried and the post disappeared. You can search in youtube video called "Craig Mullins interview about the history of digital painting" and it was submitted by Rando Archive. The portion of the video I'm talking about is around 18 minutes where Bobby asks about technology.
@maboroshiiro yeah, those people are annoying i dont respond to them at all thats why alot of art is derivative because of brush beggars and the copying and imitation of others works asking for your process instead doing there own thing
When I first clicked "CREATE", I stopped and reflected.... I remain eternally grateful for every artist that has spent their energy, heart and soul to CREATE ART that has now been stolen. AI now reaches into the barrel like an ungrateful, ignorant child poised to create another pail, lifeless cluster of pixels pretending to be authentic art.
The unfortunate reality is that this reflection will continue to be a rare experience by those who see art as a commodity rather than ART.
THANKS TIM, Great video!
Ergojosh: "How REAL artists use AI"
Tim: Stop Wasting time on AI. Thank you Tim!
Neither Josh nor Tim knows the real right answer, that's the difficulty with it. Possibly ignoring it is dangerous to your career, maybe using it too much is what'll stunt you. I suppose everyone just wants to be told what they want to hear.
@@Balloonbot How the f*ck is it dangerous to the career? If I can draw it without AI involved thanks to my skills then I'm good. And using AI would be a waste of time that could have been spent on honing my skill
@@dailydoodle42 I take "ignoring it" as not paying attention and being uninformed of what AI is doing. Though if balloonbot means "ignoring it" as "not using it" then I absolutely agree with you. It's always going to be important to pay attention and be informed of GenAI's tech advances in order to understand what dangers it poses for artists and everyone else, doing so otherwise I think is irresponsible even if it's a dystopian black hole that sucks the joy out of life on a daily basis. But 'adapting' to GenAI by learning how to use it is just shooting yourself and your peers in the foot by validating its existence.
I used AI for a background in my comic to speed up my workflow. I already know how to draw backgrounds. I don't see the issue with using AI to speed things up. Some people bash on AI for the sake of bashing AI.
@@BalloonbotFacts!
Wow people are really upset about AI. I spent 3 hours on my game art. I could have used ai to generate it in 2 seconds. Why did I go organic and draw everything by hand because I love it. Because my art 8s imperfect and that what makes it cool.
i'd take imperfect art in a game - even scribbles - 10 times over AI generated slop
A bit longer than 2 seconds and would have needed cleanup. So maybe 30 minutes or an hour.
But generating in 2 seconds is how you get slop. What you don't get is you can spend more time on it to make it look good and different. You're just arguing against low effort, which nobody argues for.
@@LutraLovegood and even then it will lack original vision. It will never match the artists imagination because it has to draw from existing material and the random element really screws with the outcome
Then you can image to video your own art. :)
I genuinely considered the technology as being potentially useful, but at every turn I just realized how much faster I could do things myself. This conclusion has just made me more confident in being an artist, valuing the stuff I like more like stylized line art and construction to the point where I suddenly do feel like have a unique style with just graphite.
Same, I personally HATE the lifeless AI pics that all look the same, I find myself spending more time looking for "real" references
From someone who gets a peek behind the curtain... "AI" is a crock of shit in every way. I've seen an apt way of describing it: Burning down a forest to create the energy needed to generate an uncanny image of a tree. We already have AI, since before electronic computers, and that is corporations.
Tim you were my teacher at cgma academy back in 2012 i was such a noob at drawing but you teach me with patience and clarity, i am so thankfull to find you again in this platfprm, so happy to hear what you have to say as a real industry veteran. A big hug from Ecuador.
Really appreciate the full discussion here, we have differing views regarding Web 3 and use of AI at all and the like- I do appreciate that you’ve tested the tools to put the claims to the test in order to prove that they’re ultimately lacking -
but regarding the actual benefits of having the skills internally, seeing students trying to circumvent the skill building, and coping with the wall in front of you, and AI being a way of getting there, it’s 10,000% my read on the situation as well. The needs that come from visual storytelling, posing, angles, hard and soft surface objects, push you to improve the inherent skills required. In the case of this particular video, some simple study of hair and hands would have gone far further than AI reference could. But when the majority of the work is typically heavily referenced, and there isn’t time in-between pinup illustrations spent on the vegetables of gesture and construction. It’s alluring to have the appearance of being an artist to others, but it’s far different to actually live the life and do the work that an artist does. Professionals have to figure these things out because the standards they have to meet are far higher, and specific. Spending 111 hours on a spot illustration or trading card would quickly lead to bankruptcy, and likely frustration from your Art Director. I know what it’s like to be making things FOR RUclips and an online audience, and I’m by no means the most skilled artist out there, but I’m also grateful that my career in the industry has provided plenty of hard knocks and reality checks that have fostered growth.
There is a basic point here that has to be said that many 'artists' out there already steal without any compunction. I make comics, and we had a variant submitted that was a direct trace of a famous artists work. If we published it we legit could have been in some kind of legal trouble were the owners of the art to press charges. "Trick" based art is so prevalent on social media that people think they can amass a series of tricks and become legitimate artists. And of course, non-artists have no clue. My writer is often completely fooled by artists who have NO foundation whatsoever and heavily rely on crutches even though he would consider himself to have an educated eye. I'm happy to see 'artists' who have an ends based approach without curiosity for the craft of art be atrophied from the field.
On the practical side however, you make a point that in order to avoid being replaced artists should learn fundamentals. I 100% agree, but I will say also another practical thing is that the point you make about how it's "not the vision they had for the piece" doesn't matter to most people. People consume art in various methods, they don't consider it very much at all and will not care about whether something was made in accordance with art direction or not. For established artists who have a style and the technical capacity this won't matter. But there will be a hollowing out of the arts because there is a suppressed demand for low and medium tier artists. Why hire a bad artist to make something that isnt good when I can get good enough and very near professional with ai. As a result people who could have been great artists but needed a runway might not see their talents blossom. That's my current apprehension about the long term effect on the arts.
I'm pretty pro AI and use AI quite regularly is all of its forms, but I HEAVILY agree with your assessment of the effect AI will have on, not just art, but a lot of fields. AI is really good at doing the bottom barrel tasks. You want a quick render of your D&D character? done. Yo want a basic webpage HTML outline? Done. A lot of the tasks that entry level and even some mid level skill jobs did, AI can do. This is going to create a skill gap where employers will be expecting skilled workers to be able to come into the field at mid level or higher, and the vast amount of entry level skilled workers without a path to practice their craft to get up to the mid skill level. And soooo many employers lack the foresight to see that the short term gains of switching to AI will be offset in 10 years when the talent pool dries up. They are literally putting all of their eggs in the basket that AI will continue to develop at an exponential rate and be able to do the mid to high level jobs in 10 years. And if they are wrong, the entire infrastructure collapses as the high skill level jobs retire out and and there is no one left to fill the gaps at an affordable rate.
@@miclowgunman1987 It's an effect I don't see being talked about enough. Many want to make moral and legal arguments, but that entirely relies on your ability to enforce your morality or legal defense. Most artists are completely powerless, even industry leaders probably. So when artists complain about art theft i say...yes and? This is already a thing. People trace, or literally use the same images and most artists have no capacity to defend their works.
I see this particularly because I definitely needed the runway. I needed to limply crawl before I could run. Someone had to take a risk on me, develop my skills and still make a profit. Who will be taking such risks in the future? If today it's 80% of creatives, but tomorrow its 70, then eventually down to 25% who only hire the best humans. Not a winning trend if you enjoy novel creation.
This reeks of infringement and not what it sounds to be--just another innovation--through the abuse of technology. The results cannot be extrapolated to other groups without the usury of infringement, to ignore everything we know is unreasonable. I see the loophole--just another innovation it isn't but covert infringement it is--we'd give way to new innovation but that's not what we're seeing here when it is a breach [edit: of] well defined boundaries commonplace in everyones hearts and minds [edit: even, or, how business is done. how people do the business [with one another.]]
It's a bit of a proletariat argument, but someone needs to build the houses for artists to live in, someone needs to make them lattes and bagels, someone has to build their plumbing. AI has ruined the idea of mundane jobs being replaced by coming after artists, but please let's also understand that while we'd all wish to see more genuine, productive (original), well developed artists, the notion that someone should be sponsoring them minimizes the labour of others.
I'm not a fan of AI slop and the obvious ways AI is being used, but I still believe that many people who could not afford exposure to low quality entertainment, education, support and tutorials can do it now. Saying that it is low quality as an argument ignores the fact that many only have terrible quality available to them now.
using AI as a developing artist seems like a great way to avoid developing actual artistic skills, rendering abilities, the ability to observe and understand space, composition techniques, visual vocabulary, or any of the other things that allow a person to create images. most of the folks I hear talking up AI art seem to either be the sort that want images quickly without having to wait for them to be created (or you know, pay for them), don't really care how the art looks, or they literally can't tell the difference between well-rendered things and poorly rendered things. >:(
You'd be surprise to see how creative some people who use AI art are.
@@h.vallis6934 If you need AI to solve creative problems, you're not being creative. You come across as someone who has no understanding of their own creative process, or the value and joy to be found in difficult work. Art is not about the end result, it is about how you come to understand your own inner life through the tedious bits of the work that must be done to arrive at an end result. And no, writing prompts doesn't count...none of the output you get from generative imagery software reflects a decision you made, an experience you had, or a problem you had to question. You tune your prompt until you find something that is aesthetically pleasing, and that will always leave you with a Product that lacks authenticity. Just start drawing, dipshit.
How many times have you drawn a thing, and you absolutely KNOW that it is garbage, but everyone around you tells you that it's amazing. like 90% of AI art users want a quick picture of the thing they are thinking about. Pikachu riding a dog? done. a D&D character portrait. Done. These things don't need to be perfect and will likely be discarded shortly after. They are also very likely to have never paid for that image in the first place. Its quite literally giving people access to art(the commodity) that they want without all the red tape that comes with finding and commissioning(paying) an artist. They are both the creator and the consumer, and if the consumer doesn't care about the quality, why should the creator.
That being said, all the things you have listed are absolutely required if you want to make GOOD art with AI. You more take the role of an editor, sifting through the many submissions the AI gives to try to find the proper lighting, coloring, and spatial proportions. The more advanced AI platform even allow you to directly choose these things through inputs of styles, 3d models, and other controlNet inputs. The better the artist is, the better they will be at controlling the AI to refine its outputs and in choosing a final product. Its akin to a camera: everyone can take a picture, but every picture isn't a masterpiece. Just because you can point and click, doesn't mean you don't need an intimate knowledge of perspective, lighting, colors, focus, and lenses, but technology can bridge that gap a good bit.
This is why people who aren't artists think generated images can be used to "speed up" making art by automatically doing the "boring stuff" with the push of a button. Like generating a bunch of starting ideas. People who refuse to learn how to create things never come to understand: the initial experiments are a necessary part of the process. You HAVE to make those yourself to warm up and begin subconscious thought processes that lead to the final creation.
Even looking at photographic references is not the same - yes, those are existing starting points. But anyone who illustrates takes a photo reference and re-draws it themselves, to experiment with what they are seeing, pick and choose aspects to pull out of it.
There could be an argument that the generated image is "just a convenient replacement for photographic references". Except photographs of reality don't distort it. That's the point of using photographs as a starting point before the artist begins abstraction. A photograph of a real lake is infinitely more valuable than an AI generated images of a lake, which is just a copy of a million distorted copies. And don't even get started on a AI generated image of a person which is mean to be "photo real". Starting with that as your reference for rendering the human image in art is a terrible idea.
What about fantasy? Things that don't exist? Here's the thing. Every piece of art depicting a fictional thing is based on real things. When you paint a dragon, you reference real animals and real biology to understand what you are stitching together to create the fantastical creature. A photograph of a real crocodile is more valuable than an AI generated array of "dragons".
@@bluedotdinosaurHonestly? Damn right.
You nailed the main problem, AI is a way of cheaply commodifying art. The main issue for me is that in the long run, a lot of art forms will lose their value.
Sure, if your only metric for the value of art is how much money it can make. An art forager who can fool museum curators also calls the monetary value of art into question. Our collective visual culture is already largely defined by corporate work and people who want to cut corners. There's something deeper at the heart of all creative work that persists and is more important than the consumerist view of art.
are we just going to ignore the fact that Fievrr was a thing long before AI became decent? Art has always existed as both a commodity and cultural expression. Art as a commodity has been racing to the bottom since automation became the norm in factories and digital art made making art cheap.
This happened to music.
@@plaguehaven but AI music is soulless. AI can only learn with really big sample datasets of a particular genre. At least for the foreseeable future, it won't be able to reproduce the soul of an artist
That's why you need to adapt and got on this train or you will be left behind. Ai is simply a tool. Use it.
Though I'm just starting out and trying to build a professional career, I'm pretty much aligned with you on all points here. For the most part, I hate over-complicating my process, it sucks all the joy out of it, and though I rely on photo and 3D reference for very specific things (mainly historically accurate airplanes, which I draw a lot of, anything else I do mostly from imagination), I'm one of those people who shoot my own references (3D or physical models) because I find my creative process limited by trying to find the perfect reference, it drives me mad for the simple reason that it just never quite matches the vision. Very much enjoyed this take, one of the most grounded ones I've heard on this topic. Much appreciated.
It’s funny that you mentioned people sketching cars by making a 3D model and tracing it. Here I am, an aspiring automotive designer, having always strongly believed that 2D pencil sketches is where conceptual design should start and be refined. When I look at others conceptualizing designs in 3D, I feel as though I am falling behind the times, focusing on the wrong thing when it used to be the other way around. You’re the first person in a long time to say that 2D sketching should be the foundation and where you should develop your skills. This needs to be practiced by everyone who wants to be a designer or artist because that’s where you learn to create emotion in your lines, designs and art.
artists are not designers and designers are not artists
@@DuckieMcduck I disagree. Designers need to be good artists, specially if you’re dealing with shapes and aesthetics.
@@DuckieMcduck Not all artists are designers but designers are most certainly artists... it takes artistry and the result of the process is art. Art can be something as simple as moving a rock to be in an interesting place and telling somebody to look at what you have done.
@@DuckieMcducktell that to the character designers.
@@DuckieMcduck you're mistaking artists for illustrators. Both them and designers are artists.
the biggest thing is that in many jurisdictions the copyrightability of the product of AI is not clear, but in the US the copyright office has been clear that the product of AI work is public domain. If you're incorporating public domain content directly into your work through "AI tools" there will be a question of disclosure because how can you sell that public domain work without letting the client know they don't own parts of the image? How useful will your character design or spaceships be if they are public domain and everyone can use them? A lot of risk involved in AI - much more than in 3d.
it's equivalent to saying cowboy bebop isn't original because it uses bruce lee's movments for the mains character. I mean, you can totaly uses free and public domain art to creat new and original artworks. No art comes for nowere, even drawing for imaination, it is a blend of a lot of things you already seen in the real world, natural things or other people's artwork.
What you’re describing would be quite literally impossible to verify.
Also, Marvel uses generative Ai. They don’t have any issues owning the films and TV they produce.
The idea that Pixar and Transformers movies using Ai to do all their inbetweens will be unable to be copyrighted is just an absurdity, it’s a pipe dream that craftsmen cling to.
The reality is, MOST media we see now and into the future have some measure of Ai integration. It’s not relevant to copyright. The few instances we’ve seen have been when people have EXPRESSLY and PURPOSELY tried to test the copyright system. No one need ever do that. It’s no one’s business how something is made.
The reason that chimp photo isn’t copyrighted is SOLELY because the guy sold the image AS “a chimp takes its own photo.” Had he never mentioned that, it would have been able to be copyrighted like anything else.
@@WhatDoesEvilMeanyou totally misunderstand. Disney can use AI within a larger work & still have copyright, but that’s not the same as using generative AI to spit out a single illustration. If you do that, Disney still owns the character but what you’ve just created is public domain. That’s why when you freelance for them now you’ve got to sign promising you won’t use generative AI.
That's exactly what I've been asking this whole time, but seems that some tech corporation don't care to use generative AI into their workflow. Is STRESSING the amount of ads and video commercials that I've seen made entirely with generative AI, and I can't understand how are they supposed to represent their products or works if they CAN'T COPYRIGHT IT
@@richardcoelho I think you have to research what copyright means, my guy...
This is the kind of conversation we need about ai right now. Really cool to hear your thoughts about it. The "don't let the tail wag the dog" is a great way of putting it. Real creative work should be guided by an artist's direction, and if you use ai incorrectly it definitely can derail your vision.
There is nothing that A.I. can add to my process. I love every step of creating my comic books from a blank page to a fully lettered page, I love every single step of the process. Sometimes I look at a finished page and I can't believe I did all of it.
I know that feeling
I think this is why AI bros don’t understand what creativity is. They haven’t felt that dopamine rush you get when you’re truly happy with the work you’ve made.
@@arlensavage4881 Exactly, very much this. I get the dopamine rush through all the process as well, as most of us, too. It's a win-win situation.
💯 I was working on a page till 3am yesterday and I was so bummed that it was late and I needed to sleep .
Facts, I got into art because I eventually want to be that cool guy at a coffee shop drawing a person I see, AI can't help me get there and I feel there's a ceiling to it before everyone starts looking for real artists again with actual skills vs building AI prompts
The funny thing is that you wont make any money while other artists will use AI to make their process, quicker and faster. Cry about it
The more AI becomes advanced, the more I realize it has no soul. Even when it produces intricate images of pure WOW, you can still tell it was generated by a machine, and not from a human imagination. I don't know why you can tell.
it's because AI does not make choices, or decide things, or even understand what it's doing. It has no understanding. It's parsing the aggregated data it's been trained on, sorting what in it's data matches the prompt given, weighs the percentages of how often prompted terms appear in images within it's data set, combines those things based on the percentage chance of those things appearing, and creates an image similar to those prompts, with some variance so that not every image is the same.
some would say that this is similar to the creative process, but the key difference is that person who is creating actually has a plan of what they actually want to create in mind - AI has no plan, and will never be able to create it within the current AI framework.
You can tell because it's not polished. AI is very bad at polish, that's why you need an AI editor.
AI art has won some art and photography competitions though.
@@bartelgrant When it was new and people were shocked and amazed by it. Most of those contest either don't let AI in now or have an AI specific category.
@@NightPhoenixPress not really man, pokemon trading cards had an illustration contest just a few months ago and tons of AI art made it within the top 100 submissions
I couldn't agree more...well said, Tim! It's still mind-boggling how many artists continue to take references from AI. We have beautifully crafted art books and incredible anatomy studies created by professional artists who have invested countless hours to provide valuable information for aspiring students. I still find myself revisiting great masters like Andrew Loomis, Norman Rockwell, and others, rather than paying for some AI site. Keep up the great content on your channel, as always!
Long videos for long drawing sessions 10/10
10/10 indeed!
ONG
In software, there's a phenomenon called Conway's law - which says, roughly, that the technology made by an organization is structured around the communication patterns within the organization: when teams collaborate on code, they write it around the organizational chart, even if not prompted to do so. Likewise, when you write code and then reuse the code later, you're "talking to your previous self", and therefore are limited by what they chose to do. And one of the problems raised by Conway's law is that you can't easily adapt the system you've built to a new context. The language you used previously informs the language you use later, which creates a cascade of effects on the software ecosystem as a whole.
Sticking to a traditional tool like a pencil keeps you completely in charge of the language, while starting from within digital positions you downstream of the effects of Conway's law: there are thousands of decisions needed before you arrive at a painting app or a 3D model renderer. When these applications were new, they were considered extremely valuable, but when examined decades on, the benefits seem more dubious: most of the features feel like "lens flares and bevel filters" - a brief fad that quickly becomes a cliche. The identity of digital art has become more like programming in nature as the number of options to configure has increased with time.
The AI tools being made now reflect a certain vision of using natural language as a command tool, with little acknowledgement of the unique properties of other languages like drawing or digital logic: those are subsumed into data. This quality has a parallel in how blockchain technologies are deployed towards a vision of tokenizing everything. Both are speculative: the language is almost certainly wrong, but has elements that could be built in a different direction. And I think that part of it is worth being excited about.
Very eloquently put. It’s true that using language prompts to generate fully finished illustrations seems doomed to fail as a result of the disconnect. Most artists will tell you that using words & making art are such different functions/ brain processes, that we don’t even like to do social media at the same time we’re painting.
It reminds me of the bb king quote how he said he’d either sing or play but didn’t do both simultaneously 😂
@@scarletsletter4466 I feel it is important to note that the strongest AI tools in the market rely heavily on other aspects other than language. The language part is just the starting line. Then you supply 3d spatial poses for the characters, style sheets for the art style, pictures for lighting and angle references, and many more inputs. Then you have the ability to draw directly on top of the generated image and have it regenerate base off of the changes you have made. Purely language based AI generation is what all of the "Non artists" are doing. Its like saying all digital photography sucks because "look at all the bad photos people are taking with their phone cameras".
wouldnt the true foundation not be drawing with your finger on a wall with your own excrement? the comical part about the whole process argument is that it suggests there's only one type of pencil, on top of a single type of "foundation" paper able to hold any artwork imaginable, which is incorrect.
you'll always go down the rabbit hole of "deciding downstream" if you wish to make "real art" in one way or another, and the most common thing to happen is being overwhelmed whether with digital or "traditional". most people dont know what some pens can do, and much also have no idea what their software can do, they find that one thing that works and stick to it. fortunately, you can't smell the digital.
Drawing is a non verbal activity- Artist's don't talk to their pencils- Drawing is the most direct and pure way to communicate a visual idea.
I suppose some very clever tech bro could invent a technology that works by typing words in order to create pictures- but that would be stupid right? Because if words really could define pictures then we would not need the pictures and the blind could see just by having the world described to them- in words.
Text to image is a genuinely funny idea because it fails at a very basic level, a failure that arises from a very human misunderstanding. As a species we are so entranced with language that we make the mistake of beleiving that words can actually capture reality.
But you can't drink the word 'water'- and you can't really create visual art by using a bunch of words to describe it. Certainly you can approximate a visual idea by describing a picture using words but this is not making art, this is at best art direction, not making art.
So the fate of the AI 'Artist' is to use words and a bunch of other tricks like 'image to image' in the hope that he can exercise some degree of control over his work- but in the end it's the machine that calls the shots, not the 'Artist.'
Has anyone noticed that AI artists use the names of human artists in their prompts- or at least they used to until the AI developers stopped them doing it. But has anyone ever heard of an AI Artist using the name of another AI Artist in their prompts? No- me neither. I wonder why that is? ( Actually I know why that is- AI Art looks more like AI Art then it does the work of any particular user of AI.)
The point is that 'Text to Image' is a really silly way to try to create visual Art, which is why AI Artists are never going to be able to deliver the granular control over their outputs that their clients will expect them to deliver. Any short term gains in 'productivity' from using AI art will in the longer term be offset by the basic intractability of trying to use words to define pictures.
As a writer, words can capture reality. A good writer can bring their writing alive in your imagination. "AI" art doesn't understand what words you are using as prompts, it doesn't understand anything. It only has key words which it scans, and spits out an image with, using a whole set of variables, loops, and conditionals that's programmed into it. Language, however, is the best conveyor of an imagination. There's a reason books have the broadest range of stories - words have no limitations.
@@magicbuns4868 I don't disagree with what you say here- words can create whole worlds of imagination. A book like Lord of the Rings for example creates a compelling mythological world that many people find deeply engaging.
But what words cannot really do is express the subtle nuance of the visual. For example; If you wanted a portrait painter to paint your likeness would you send them a written description of your face? Or would you send a photograph? In this instance words alone would not be enough- no matter how detailed a description you wrote it would still be impossible to capture your likeness in the way a photograph does- which is why Passports have photographs.
Another example might be a visit to the Louvre in Paris to see the Mona Lisa. If you arrived at the gallery to find that the Painting had been replaced by a detailed text description of the Artwork would you feel that in reading the text you would have seen the picture? Probably not. So you can see that the idea of using words to precisely define images is not really viable.
Going back to Lord of the Rings- before the movies came out- eveyone who read the book had their own image of what Frodo or Gandalf looked like- your Gandalf, the one you pictured in your mind, probably looked different to mine. In fact each reader of the Lord of Rings had their own unique understanding of what the world of Middle Earth looked like. This is one of the great virtues of the written word- we can each have our own unique vision of the worlds of the story, precisely because the words allow us the freedom to imagine those worlds in our own way.
But this lack of precision does make the idea of 'Text to Image' a bit silly, because if words cannot really define in detail what an image actually looks like then using words as a way to control an image making machine is inherently nonsensical. Those who claim that the images they get from their AI's are 'exactly' what they had in their mind are really engaging in a form of self deception in my view- they are in effect retrofitting their expectations to whatever the AI gives them- like getting a pair of socks for christmas then claiming that socks was exactly what you wanted all along!
In this sense AI Art is a bit of a con- no one can really control what these machines produce with any real precision- the best you can do is get an approximation of your intent, filtered through the AI's training data and it's programming and further limited by the fact that no amount of words can really express in detail the image you are trying to create.
AI Art works fine in situations where all you need is a crude approximation of your intent- but as soon as you want any real artisitc control they become more and more frustrating and the limits of a UI based primarily on words become ever more clear.
The ethical/moral/legal (and environmental) issues are obviously very important, but this message is critical too. Even for people who can manage to convince themselves that those issues aren't important, or that they're somehow enlightened and different, I don't understand why anyone would want to stunt their progress and shoot themselves in the foot with this kind of thing. Even if you never take it to the professional level, there's such joy to be found in exploring ideas and making choices, in learning how to learn and in learning about yourself and your style. In learning about your contemporaries as well as all who came before and in their process and choices. In filling pages with thumbnails! If people want to rob themselves of that growth, experience (and possibly community as well), that's their prerogative, but I do hope they understand the massive tradeoff they're making. Hopefully messages like yours will help inform.
I loved your take on this. I have struggled this past years not because of AI, but because I was in a bad state physically and emotionally. I was kind of burnout. So I had to stop, think about myself, about what I wanted, about the technology emerging, aka AI. And I arrived at pretty much the same conclusions. I always wanted to draw whatever I wanted, it was not about the end result, not even about me expressing myself, or my vision. But it was more about me being able to do it, If that makes sense. But many tools take away that feeling of "I did this, like yeah I completely did it" So I started working on my fundamentals again. And I understood that AI is designed to be very easy to use, even for normies. So in the case that the use of AI becomes normalized and accepted because all legal and ethical frameworks were properly stablished. I would be able to learn it really easily (I mean, the tech was like two months old and we already had AI "artist" and prompt "engineers", which I consider a bad use of language, but that is another story) and then I would have a pretty good competitive advantage. But if I can evade using AI and be competitive. I'll do it. I mean drawing is really fun! why would I be doing it of everything that I could if that wasn't the case? So using AI kills most of the joy.
Ai took low and mid jobs, which allows artists to do their craft at the same time paying their bills. Without that you can't get grow professionaly, while AI is getting better and easier to use. Let's face it they took our art without any consiquences and now they are selling blended pieces of our community labor. AI just made hard way of life of being an artist, even harder.
AI made my art better
Because I realized how much my art looked like it ...
It forced me to do the things that I noticed A lacked , that being : story telling , expressions , interactions , composition , im sure there's more.
And yes, I use thumbnails now , great way to visualize the final piece , especially if your going for interesting color combinations
Thank you for giving nuanced takes on these subjects 😭 it’s very sobering.
I'm tired of AI.
AI-produced images are not art. They are the result of prompt-writing, nothing more.
Yes, let's call it literature.
And by the way, 30 years ago, digital art was not art either !
you snowflakes are in denial
@@h.vallis6934 Yes it was. Digital art was/is (for the most part) simply a new media to use many of the same skills which are common to any visual art: composition, form, colour, etc. It differs from traditional media obviously, but there is still creation happening and you need to have an understanding of the fundamentals to create anything. Replace one artist with another, and different results are obtained even if the means are digital. With AI-prompts on the other hand you need absolutely no artistic knowledge or skill to "produce" an image on a screen. It is just a gimmick for non-artists to generate some images which allows them to feel like they get the results of being competent in the short term without actually being competent in anything. Theoretically speaking if two people type in the same words, they will obtain the same result. There is no artistic skill involved there, only a type of curation of computer-produced garbage (and as Tim explains in his video here, this is not even worthy of the status of "direction").
@@h.vallis6934 Let's call it what it actually is...an ugly mush of stolen content.
this video is not long enough
we need a series for a playlist!!!
@@WickedPawn yeah jokes aside videos like this are super nice to listen to while drawing so I would love that ^^
I draw because I love it, so I don’t want a robot trained on stolen assets to do it for me. No thanks!
The problem for me is not relying on reference, but Josh over-relies on it. Like, he needed to pick a reference to draw an elf ear. Its not rocket science, its just a pointy ear.. also the hair on his illustration doesnt make sense, theres hair strands spawning out of nowhere, and that is the focus point of the image, very noticeable. doesnt look like he really took control of the 'tool'.
As an Artist myself, who started out with a Pencil and Paper in the 90s (now a 3D Artist in games) I completely agree with everything you've said here. Cheers! :-)
"The older I become, the more I realize that drawing is the most important of all the problems of picture making."
Joaquin Sorolla - Spanish Impressionist.
It's a good point that if you learn how to just do it, then everything else is just a tool to get you there if you need it.
Horrible ethics aside...
In most cases I dislike the idea of using AI for reference because it's going to be a distorted version, and then it just runs into the copy of a copy of a copy problem of decreasing accuracy.
I always felt like the most promising future generated imagery had was basically custom tailored entertainment. It's like a toy, not the same as seeing an artist's work which is a form of human communication. That's where I see a lot of conflict is in confusing those two things.
Wow you hit the nail on the head.
The main thing I saw from the video from the artist you’re referring to is there was a big lack of having a simpler process along with not understanding that their issues were all foundational.
It reminds me of something Will Weston said, that the problem a lot of artists have is that their methods are too cumbersome, and because many of those artists don’t bother learning how to simplify their methods and apply the fundamentals in a more simplified way (like a 20th century illustrator would approach it) they tend to compromise on their overall vision.
I tried creating art with AI when this craze was just starting about 2 years ago. Generated a bunch of sci-fi robot images then painted over them heavily. It ended up looking like something I would've painted myself. But the entire process felt so boring, there was nothing rewarding about creating that piece of art. Proving to me that sketching a bunch of ideas and working on them slowly is the best part of the creative process. "Maybe people can use it to create thumbnails or something" I thought, and that would be it. Never expected people writing full names of artists as prompts and generate uninspired versions of their beautiful artwork, basically stealing their years of experience.
I miss the art scene that existed before GenAi... we were so excited to make art...
i started learning to draw because i wanted to have the physical ability to create something with my own hands… ai takes that away…
This is one of the best video on the topics of ai to reassure those still optimistic enough to study art. Thank-you good men
This whole "drawing from imagination" certainly is a good exercise, but it's not what everyone strives for. You have to draw from reference in order to eventually draw from imagination. But when I say "draw from reference", I don't just mean draw what you see, but rather, understand what you see. Try to challenge yourself, like, what would the pose look like in a different perspective etc. However, I have to note here that the more stylized the artstyle is, the more simplified the forms are and thus the easier it is to draw things in perspective. Meanwhile others want a semi-realistic artstyle, which requires more knowledge about the human anatomy etc. It can take years and years of drawing in order to get to that point and references can make that easier for you. I still take photos of my hands if I'm not sure whether the hand looks right. I still look up references of arms and legs in specific positions cause I'm not sure if I forshortened them correctly. But what I agree on with you, is that unless they wanna draw realistic or copy an actuall reference, people should stop wasting their time finding a person that does the exact pose they wanna draw. It's not helpful, you will forever rely on references if you don't dare to try and challenge your imagination every now and again. At least in my opinion, but maybe others have good reasons why they continuously draw from existing pose references etc.
I agree. I think referencing is so important for artists to utilize properly. We really ought to move past shaming artists from using other images and photos as guides to better enhance their artwork. Drawing from imagination is largely just referencing from memory. The more reference you use overtime the more your visual library grows, allowing you to draw from imagination more reliably.
tbh, this video was something I needed. I'm someone who gets stuck in the same mindset and continuously worry about it. Same thing happened when I heard about Ai and artist. As an art student myself, I can't see myself in a different career path, so I started to panic. But somehow this video help me calm down from that mindset. Thank you
This was extremely relatable to me. Currently, I am an incoming junior into an entertainment design program and in the last year or so I have completely pivoted from illustration to 3d modeling. I can't deny that the programs limitations change what type of art I am creating. This video definitely helped me see what is truly valuable when creating art and I am glad im not the only one getting stuck on this.
Thanks for this really enjoyable video Tim, I really appreciate the balanced view of AI as it currently stands. I'm 7 months into my first design job inside a small company where it is obvious there is a fascination with AI from a business stand point. As you have said, if the AI could do my job - it would and I wouldn't be working but that still doesn't entirely calm my nerves about my employment in the future. I feel like I will need to try and expand my artistic incomes just so i dont have all my eggs in one basket
Saying that using a reference is the same as using AI, is like saying that using a recipe for a dish is the same as asking for someone to make it for you.
Not at all.
Using a recipe is like following a tutorial.
Using AI and references is like... watching somebody cook next to you, doing the same and saying you came up with the recipe lol
If you copy a reference exactly and say you came up with the image, that's just wrong mentality. I don't think a lot of people do that.
References are used best for learning specifics, be it painting styles, lighting, colors, etc.
Using images of a sunset to learn how lighting in sunsets work, that's following a recipe for sunsets, albeit in a more abstract manner.
The longer I watched this video, the better it got. The whole point of better foundations really hit home.
Great video! I agree - I'm just tired of even hearing about AI art. It's not going to stop me from doing my own art, so I don't really care about it. I just wish it was easier to filter it out online, especially when searching for reference images. I don't want to be accidentally using AI art images as drawing references without realising it.
10:45 so I am a game developer and I do think LLMs are pretty great at bug fixing or writing code. The thing is, it isnt sustainable to make an entire game on its own. Its sort of like, you have something that can design machine parts, but it cant design machines exactly. So you can get small components basically for free, but you then have to understand how those components work in order to make the actual machine. And heck even to make the components to the right specifications, you need to know what those are.
You are best off just doing as much as you can on your own, then trying an AI when you hit road blocks, if the AI cant fix your issue, ask someone on a discord.
Its basically a replacement for stack overflow more than anything.
Really I have noticed the same thing in programing as you have with AI art. I can waste a lot of freaking time if I try to use it for the wrong thing. I'd rather just use it for simple syntax or catching a programming error, and actually program everything myself.
One of the most unique and thought provoking takes I've seen out of this whole situation. I'm glad you shared it.
one of the most necessary videos about making art
The Director comments. I have heard successful Directors say, some days, some shoots, some scenes - they have no idea how it's going to work out. Sometimes it comes together in post, when they start to assemble. In fact, I've heard - from a storytelling point of view - entire movies have been rescued during the editing process.
That’s a way of humble bragging or saying they direct by instincts. They’re still a human who is making decisions with their brain. Even if we take those comments literally to mean they truly didn’t consciously “know” why they made decisions the way they did… there had to be some reason. Even if they were using a random number generator to come up with ideas, that’d still be a method
@@scarletsletter4466 Yeah, I think sometimes the creative process finds itself, in a way. Sometimes. I'm sure there are days when a strong vision from the start is followed through, right to the end, and other times it's more like play, where you're kinda just figuring it out as you go, following your instincts, fixing it in post, and experiencing the magic of 'happy accidents'. It can happen - and I've heard, directors, song writers, et cetera - talk about it. I've experienced it. I even heard a video game developer who worked on Half-life 2 describe the game coming together in a similar way. That was surprising to hear, but given how long Valve can take on a project, maybe that partly explains why. I'm just saying I don't know if I completely agree with Tim's take on the Director thing.
As a 3D artist, I use enhancing AI to upscale renders, upframe animations, or denoise images. That all leads to less render time needed to achieve my end result.
When it comes to generative AI, I am basicly agaisnt it in all its forms. I mean, if somebody uses generative fill in photoshop like an advanced copy stamp, only generating meaningless background, it's lazy but if the hero object in the image is crafted without AI, I wouldn´t even notice the difference. ANd maybe it´s not even lazy in an age where we all need to be fast in execution.
But since generative AI is used mainly to completely circumvent the progress of coming up with a visual concept and then flesh out the detals in painstaking work, I think it is just killing the need for skills. Funny, since the emergence of generative AI I have grabbed my pencil more often than in the last decade. I was focussing of 3D modeling stuff but that stable diffusion stuff really reminded me of the fact that I have neglected my drawing skills even before all that Ai stuff. I took it as a waking call for me.
I really needed this video, Thank You! I feel like I need to pay you a therapy session bill for this! 😅 I'm constantly anxious about AI as a freelance illustrator, and this video has helped me so much Tim, Thank you!💛💛🙏
Oh, and your Art is incredible, I think I need to purchase one of your courses
There are entire worlds trapped inside a pencil, pen or brush. It's up to us to let them out.
AI is not an achievement. It's buying food from a restaurant, putting it on your own plates, and telling your guests you made it yourself.
"You know, these hamburgers are quite similar to the ones they have at Krusty Burger."
"Hohoho, no! Patented Skinner Burgers! Old family recipe."
Restaurant? My facebook feed looks like the dumpster they throw all their leftovers into.
@@T61APL89 Facebook is full of bots, so no surprise there
but also stolen lol
I've always thought AI would only help artists stop improving and you kinda touched on that. I don't think it's harsh, it's just the truth, you stop trying to improve your weakest spot, you won't get magically better.
I never understood the claim that people who don't use AI will be left behind, either. Like we can't all learn to type a couple words in a couple minutes of messing with the stuff? People felt so pressured into it when it started to pick up attention, for fear of losing work, but it's made to be easy to use.
Just discovered this channel and decided checked out your online page. I love to draw and recently started exploring 2d animation and doing some fantasy writing. That being said, I’ve never really gotten past the hobbyist phase of my artistic journey as I struggle a lot with anxiety when it comes to my art. The moment you mentioned Asterix, Obelix, old video game instruction manuals, and Princess Mononoke on your ‘About’ page, I knew I was in the right place to get some guidance with my craft as those were all discoveries that really influenced my love for art through my high school and college years. Watching your free Mini Workshop now, and I’ve just subscribed! 🔥👍🏽🔥
Haven't watched the video yet, but that work!!! Omg, i love it, probably my favourite out of your recent ones!
I love long talking videos and your voice is very nice to listen to - also I am an artist and i enjoyed the vid! I'm not really interested in AI, but your video did solidify my own feelings and helped me clarify the way I communicate with others about art and AI in general. I'm a big proponent of using the simplest tools as creatively as I can :)
Thank you!
Thanks Tim, I really needed to hear this today. I keep going from a doomsday scenario losing my job, to feeling like things will be okay. It's taking up way more brain capacity than I'd like, but I can't seem to push the negative thoughts away. I guess all I can do is carry on doing what I've been doing and hope clients still want to work with me over AI.
Yeah I think the FUD takes up a huge amount of brain space!
"Are we being luddites?" interestingly, the original luddites was a 19th century textile worker movement. these workers saw their bosses using profits generated by their work being used to buy automation machines that was going to decrease their wages and cut jobs, inxtead of creating the same level of output with less work for the same pay.
If someone calls you a luddite, they're actually calling you pro-worker, someone who advocates for the principle that new technology should benefit all of us, not just the owners. This relates to AI in that very sense. If it was a tool that works, and it could help peoples workflows, instead of it being used to "save" on labor costs and cut jobs, it could be used to let the same amount of workers create the same amount of productivity with less work and the same pay, or increase productivity for increased pay.
Capital's drive to generate profit regardless of the cost is what creates this tension between workers and new technology, something our predecessors the luddites understood.
Yeah... I'm aware of all that. It's an interesting story. I think the ship has sailed on the usage of the term though :) the winners write the history books. It's not a good look no matter how you 'well actually' the verbiage. We are best off learning from history to have another crack at standing up for pro-workers rights. It didn't end well for the Luddites...
Here after the ergoJosh storm. Finally someone who speaks knowledge and truth. Thanks for being a real Artist Tim.
Thank you Tim!, I’ll keep creating and do what we can and develop by my own skill step by step. Not ai 🤖
People don't realize the release of Endorphins while making art with your own bare hands and at the same time building a skill that will most likely be gone in few years so that Art once again will be done by the elites and sold in closed off auctions.
You underestimate the endorphins people feel while being able to physically manifest the thing they were just thinking about into physical form just by typing its description into a box. There is a reason millions of people are subscribing to Midjourney, and millions more have downloaded Stable diffusion. And AI doesnt take away anything. Just as you still find people drawing and painting, even though digital art is easier and cheaper. The only thing that will shift is art as a commodity, not art as a form of expression.
@@miclowgunman1987 I’ve seen a lot stories of former ai bros who admit they stopped to enjoy “making” ai “art” even though they were getting results they wanted. It’s just not fulfilling in a long term. In my opinion, when you generate a picture, even if it’s everything you wanted it to be, it is more like randomly finding preexisting image on the internet, because you didn’t do anything except typing some words lol.
I'm sooo glad I learned how to draw. I can easily draw whatever I see in my imagination. Sucks to those that need AI or 3D to even start lol
Great video Tim! Fully agree with everything you said about ai. It is just way too hard to get anything useful out of it. Having solid foundations and do everything yourself is the way to go. Using 3d on the other hand is super usefeul and changed the whole industry drastically as you said. The cool thing is that you can design and iterate pretty easily in 3d these days, so it is just a tool but not a crutch. You still need to be purposeful in how you use that tool. Ai is just too random to be useful
This was a really thoughtful, measured consideration of our current situation. Love the focus on the need for the fundamentals/hard work.
Thank you for being the voice of reason here!! 🙏 With the current takes of some bigger artists, I felt like even artists are forgetting how these genAI programs use stolen artwork 🤐 For me, the lack of usefulness to artists is secondary but seeing you speak out against AI fills me with hope regardless and I'm super thankful to have found your video! I hope even more people get to see this 🙏🥳✨
As someone who has contracted artists in the past, I choose people who have a consistent personal style that I like. Nowadays, when I see a portfolio from someone with different styles (and especially varied quality), I tend to assume the rely too much on AI art (or that the crappy drawings were them without AI, and that the suspiciously highly rendered images are AI doing most of the heavy lifting). That's a turn off for me.
The pinnacle of working with an artist for me is chatting with them about the basics that make a character or scene identifiable and the impact I want the image to have... but let them create the scene that best show cases their style which I already trust.
I don't have to be sold on the benefits of working with an artist over AI. I just have trust their art style and by sold on the quality of their communication/collaboration.
Definitely listening to this while sketching. ❤
I love your style! Somehow someway I have never heard of you and you are so awesome and so is your channel! How have I never heard of you?? I need to buy your sketchbook asap!
The Silicon Vally hype culture.... hit it right on the head. People spew LOTS of nonsense especially in the film business right now because there's a lot of transition and things feel kind of dark right now.
I had a lot of fun messing around with Midjourney in the beginning before realizing I'd rather get back to drawing and that very few use cases of importance to me would be better served by using AI. I'm not ideological about it though, and I think there people making cool work with AI, usually by screwing it up in some way to create strange new images.
Nothing can make an artist happier than a piece of paper and a pencil.
loved the video! im so glad you talked about the importance of learning the foundational skills and how vital it is and allways will be for a creative person, its what i told myself when i started to worry about AI emerging in art, and its so funny if you actually play around with A.I as an artist who understands how to construct an image from scratch using fundamentals how frustrating and boring it is to use and you end up thinking it would have been way quicker just to sketch it out. i myself use alot of 3d in my professional work when it comes to ships and vehicles but before i learnt 3d i knew how to illustrate and design a vehicle in perspective by hand also so im not 100% reliant on 3d software, just using it for speed when it comes to clients demands, i still sketch and paint freehand for practice/fun and jobs where i know i have enough time to do it all by hand.Again i think its so important to be grounded in the fundamentals and everything you said about that fact and parts about being a true director with a solid idea was spot on! glad i found your channel
I click this video expecting another "is better if you don't bc AI will take my money away" but im glad to find a real artist just talking about how is boring and lacks quality.
As a 3D artist (that has drawn a lot too, which is harder if you ask me lol)I use it to TRY and get an idea out of my brain with it, but i fails more than succeeds most of the time, also for human reference is absolute trash, unless you want to draw the same girl with questionable anatomy 1000 times lol
I have been doing this for almost a decade and some, and honestly AI is a fun curiosity, is kinda fun to get results from text but thats it, if you have a visual style you mostly will end up doing it yourself bc AI just makes it feel icky idk what it is but is very cliche all the time and since you didnt make it you cant iterate anything from it. Not taking into account it takes A LOT of tries to get barely close to what you want, so in speed I lack at least.
I can see it making it faster at some point but if you have a quick mind with a good visual library drawings win 10/10, maybe some cheap cheap businesses can use it to sell pens or something but on the game industry and real good art like 2D animated movies, I dont see it making it too far, taking also in consideration some ethical issues, for now I used it to make fun AIs of my dog and some profile pics to avoid data scraping of my face lol, but I will try to make some content soon so my face will be all over the internet so not even that now lol.
Anyway good video, you seem like a very passionate artist that enjoys doing it, which imho is the most important part why would you want to skip the fun part of life
To each his own bro!
I 3D model stuff that I can't draw in 2D and then pose it in the way I want so the perspective and stuff matches with everything else etc
I miss the art scene that existed before meatbags started to tape bananas to the wall and call it "art".
If I had to make a video/talk about AI it would be the very same as what you've just done here.
I am a professional illustrator, over 20 years of experience in various organizations and now, like you, in web3 and blockchain, because, well, we need to make a living.
On the side I've always worked on my personal artwork and more interested in the history of fine art than gaming. Great passion for comics in my early years, though.
My organization is pushing for AI more and more to the point it frustrates me and my colleagues. We see what they are trying to do. We use Midjourney at times to come up with references or sometimes even use its generations in the final, with due modifications. You can imagine how that makes me feel.
First of all, AI never makes what I have in my head. It can be useful in industries with no artistic culture and seeking very standard aesthetics (most of them) and imitations of other things they see. In that case AI makes exactly that and it can be a shortcut. But when I work on my own pieces I am in a totally different mental zone. I came to feel when my drawing and sketches come alive and resonate with what I want to say. That "aliveness" is something that only those who dedicated years to a craft can understand, certainly not marketing managers or similar characters. But when an artwork goes through such a process then it does transcend and speak more.
AI removes the process. That's why it can't be considered an artistic tool for illustrators and designers. By using AI you stop learning what art is, how to make it, how to feel it.
AI can be ok to substitute non-artistic processes, like perhaps some tedious manual tasks of 3d production (standard rigging comes to mind perhaps), but not much else.
My solution to all this is: if you are an artist, keep learning the fundamentals but also dive deep into where your heart is.
Before using a reference, try to draw that thing as best as you can, until you bleed on the paper and only THEN find a reference and compare the two. Only that way you can internalise the process and strengthen your drawing skills.
Don't fear mistakes, because mistakes can be inspiration.
Imagine you live before the internet era and work as if you didn't have google nor many books to refer to.
Use your memory and observation skills to create a mental library of anything is an interest to you.
Let music, books, stories, people...anything really...inspire you, not just artists.
Then your work will flourish, will have more and more soul, become more unique and no AI will ever be able to beat that nor will ever be a seductive alternative. AI has been created by people who don't know art.
Art is a process. Artwork is its outcome. Live artistically.
This an excellent discussion. You've just earned yourself a new subscriber. :)
I've done some art using illustrator and photoshop, and I drew a lot when I was younger. I'll admit my drawing skills have languished over the years, but I still 100% need to start an idea by doing sketches. Even if they arent something that I would feel would sell a client on my work, but rather just work for me as part of my conceptualization process to help me bring the fuzzy concept in my head more into focus and then have those sketches as a reference to draw from while using illustrator or photoshop to create a polished, high quality, professional, and fully realized result of my initial fuzzy/out of focus idea that I first had in my head. As soon as I can afford a GPU I plan on starting to learn blender, and I still plan on using pencil and paper to help me work out things like form and layout as part of my initial process before any 3D modeling takes place.
Hi Tim, I just found your channel and I appreciate you sharing your perspective! I've been a character concept artist in VFX for almost a decade so I thought of chipping in to your thoughts.
I still think AI gen is at least a useful brainstorming tool even at this early stages. Yes, making 5 drawings by hand does build that muscle and is authentically human, but in that time an 'AI artist' can provide the client with 500+ images to choose from. Is the sheer quantity and speed of production that wins over authenticity and quality. (although image gen model are getting so good and realistic by now that only something like the top 5% of artists in the industry could compete in the value they can provide over it)
I do agree it's worrying to think that humans will progressively lose their artistic craft and way of direct expression through it if industries will rather opt for the cheapest and fastest solution which currently looks to be AI.
I also think that if the market goes in that direction, without intentional appreciation towards human made art, we'll all be at a loss from a human originality point of view.
From a different angle though, having generative models which 'digested all of human made art' that is now in the hands of everyone sounds like a powerful progressive commodity which can indeed allow everyone to design, brainstorm ideas visually and create a world for ourselves that may overall be better than for humans to rely solely on the 10-20% of the population who dedicated their lives to mastering the craft and acquiring the skills needed to realise people's visions.
I see generative AI in general as an immensely impactful technological innovation which for better or worse is causing a global multi-industry shift which can't really be stopped, but can be veered responsibly based on how we collectively carry ourselves.
Thanks for the reflected point of view - It's a great help to figure out the topic.
I appreciate that you're looking at this from an artistic perspective rather than ethical, I've been feeling like I'm going crazy because people keep using generative images to inform their art and saying it looks good. It always looks bad. At best it's boring. I've tried and never been able to get anything that looks good. Ergo Josh's hair reference looks awful. It's just a big mess and it's reflected in the final piece. The parts he did himself are nicely rendered but other than that I don't think the piece is that interesting overall. And he took literally over a hundred hours to make that. If he had just looked at real hair and practiced for those hundred hours maybe he'd be able to draw hair himself.
I'm also a veteran in the comic industry (almost 20 years), but my dream has always been to create for the French comics/bande dessinée market. How did you reach them? I have an agent, but he doesn't have contacts in the French Belgian industry. Any tip? :)
love this! I start with thumbs and concepting the characters (usually for a series of illustrations for children's books). My work is easy when I am working with simple angles, but I do tend to use interesting angles for which there are rare photorefs. So I will digitally sculpt and sometimes use 3d environments to speed things up. I'll use AI to generate photorefs that I would normally have trawled to know what a thing will look like. I am about to get a 3d printer so that I can print out my references as I am always fighting my way to work as little digitally as possible. Sometimes its just about being fast and getting the concepts cleared. I don't render colours though - everything is essentially models to work from, be it 3d or I sometimes make models. James Gurney I think is the master of making models to work from IMHO. Great work as usual!
I respect your commitment to traditional art and fundamentals, they are important, and I love your art as it is inspiring. But I’m using AI generator inside Krita to fill clean canvas with something to start with. I don’t think that AI will be able to replace artists. It could draw generic assets, but not composition, and not finished compositions, that’s an artist job to guide it. People with vision would stand up, not the tools. But for now the level of control over AI is still very limited, and that’s the problem which creates divide between artists and tech
I wonder if the quality of art has dropped over the last century is because people aren't willing to put in the work? Is that why some just scoff and say, "Anything and everything can be art!" But if that's true, then nothing is art. Is this how we go from carved marble statues that look alive to making some large abstract statue out of clay? Has art fallen so far that people say the "meaning" of art is the only important thing, and create an "art piece" of scattered trash that resembles a wild party that never got cleaned up?
Art has NO shortcuts! We mustn't get lazy, but I fear social media has affected art by forcing artists to churn out art faster than they should; "You can't rush art!" that old guy from Toy Story 2 said, and perhaps artists should heed that advice.
I cannot thank you and other artists who make videos like this enough, we need to just put our foot down and say no to AI making art for us.
hey tim I hope you see this comment, I would really like to make art my career much like you do. there are dozens of videos about different jobs in the art field but when I look I just can’t seem to find any actual work. it you be awesome if you could share how you find work or how you found work in the past. love the content keep up the great work.❤
literally one hour of simply saying "have you tried picking up the pencil and drawing?"
Completly agree wth everything though i'd say I'm even more harsh about AI as a useful tool. I don't ever see a computer replacing the human heart.
New to art as well 3d art. This video was really helpful. Thank you for making this. Almost all tutorials recommend sort out references as the first step. Using pure ref or something as the first step. And whenever I try to do that, i lose all interest in that project. Trying to use mid journey to create reference images kills the creative satisfaction completely. I always prefer using references or ai only when I am stuck with visualising something midway through my artwork, like unable to exactly imagine how does the foam around the rocks in a river looks, which is just a part of my painting, or how to get a the right lighting for a certain kind of look for a blender render. But starting with ai or a reference and trying to improve never works for me. I have been pushing myself to try to learn the reference first thing. And it was making me hate the process. Thank you for this.
IMHO most people don't care nor pursue the artistic side of an Artwork, be it digital art, architecture or concept art etc. Art is for them just stunning impactful visuals, and that is just AI-art is. A bit similar like fast food industry of art, only that Art isn't that essential like food for most.
I always found delightful to search for references among photographs. It's a way to broad your artistic view and improve your artistic taste, also learn to appreciate other people works.
Like you, Tim, I have tried Midjourney on the past (when working as a graphic designer) and found it extremely boring. I prefer to use stock images and primp them.
I've had this exact problem of "letting the tail wag the dog," but with 3d instead of AI. I consider myself a 3d artist/animator, but I've recently realized that my 3d art is just... bad lol. And of course it is! I've skipped the fundamentals! So I recently decided to finally learn how to draw in earnest, and boy do I wish I had done so sooner. I'm having the time of my life just learning the basics. I'm looking forward to returning to 3d *after* I master the foundations and develop a unique voice so I can see how the tech will enhance my work. It's become clear to me that drawing is really the best and only way to learn and master the fundamentals of art.
I spent half the video to find out if the mic was real or edited, then I gave up 😂
Great content! You just got a subscriber!!
I think people have an intuition that master the craft of art over 15 years of hard work is worth it even in the advent of a miracle box that does the work for you and it's not just the sunken cost fallacy. They are just waiting for someone to precisely articulate the reason as for why. I know that I am. I have not heard anything yet, that convinces me entirely. It is something deeper and philosophical.
also the situation is a bit more multifaceted than just the consumers nor caring, while it is true that most consumers dont care about how automated a particular industry is unless it affects them directly, the fact of the matter is that generative AI isnt just a threat to just visual artist but any job that requires expression through a visual medium, that means models, photographers, 2d artists, 3d artists, news broadcasters, actors, content creators etc, and not to mention the copyright cesspool that AI has caused with all the impending lawsuits. The fact of the matter is with all the industries that it threatens not just individual jobs, people will start rejecting this type of content in mass since its a tool at its core designed to undercut people. Eventually it will also cannibalize itself because if AI can be copyrighted while simultaneously feeding off people's work then eventually nothing will be safe within copyright law, since anyone can use your "ai work" and use it no one's work will truly be safe under copyright law. Also the huge potential for political misinformation with deepfakes and how it might affect evidence in law. SO huge rant aside, the consumers dont care for now but they will when it's a threat to almost every job. Sadly sometimes things need to get much worse before it gets better
I think directed image generation has its place in a support role. Image generation is really clunky when it comes to creative control, but for in-betweens where you mostly redraw the scene, for creating non-descript, unimportant details it could very well speed up a project.
I'm a concept artist in between jobs right now (and I realize I might not see the next job). I generated a knight that looked kinda cool in midjourney, fed that in Hailuoai that generated a 360 turntable out of it, fed that into a photogrammetry program, and it actually created a 3D shape that I can use however I want. Did the same for environments.
I think we can look at concept art in two parts that's essential of what we do for a company. We help the creative process and as a by product, create things that other departments can work off of. It takes a lot of skill to generate 2D designs, and to translate them into 3D. Concept artists were used to solve, to varying degrees, for both issues. However I just demonstrated we can skip the whole process of a concept artist entirely and head straight to cleaning an end product. Do I want to do this in the future? No. I think i'm checking out and become a night guard, draw/create in my free time and will definitely not share art online.
That sucks, sorry to hear about your job. I totally get where you are coming from.
This is not career advice... but I think the more you can focus on the actual design aspects of what a concept artist does the more you are going to secure your future as a professional artist.
Many companies and projects have essentially been cookie cutter copies of each other for the last decade or two as games have become a huge industry. This derivative ideation is no longer valuable as AI can do a good job of regurgitating concept art sludge. But the technology is fundamentally built on the logic of serving up what it thinks should go next based on it's dataset and neural net. This means that the value of this type of work has gone to zero. But there will be more and more need for actual design and innovation at big companies.
I think any large corporation leaning into AI for ideation is going to find itself easy competition.
The biggest asset a large ip based company is going to have in the future will be it's ability to come up with actual new ideas and get them to the marketplace before they are copied by cheaper AI competitors.
I think it's best to always see yourself as an artist separate from the jobs or career you do. The career will change... our role will change. but your artistic self and the stuff you are interested in should be something you continue to develop and build on.
The only advantage of getting a job in a creative industry is you get more practice to build your skills.
Thank you for sharing your thoughts about this. Artists bring originality where artificial intelligence generated images are recycling what is already existing.