AI is GOOD THING for 3D Artists - Here's Why....
HTML-код
- Опубликовано: 3 июн 2024
- In this video, I offer my perspectives on the topic of AI and the impacts it has (or doesn't have) on the lives of professional CGI / 3D artists. It's okay if you disagree with my views, but make sure to be respectful about it if you decide to leave a comment below.
Развлечения
It seems the impact is becoming more apparent. Let me share an example: Earlier this year, we had a project where we initially planned to provide style frames and storyboarding for a client. We estimated the cost at around $12,000 for four senior artists to deliver high-quality concept art and storyboarding. However, the client opted out of this particular preproduction scope and instead utilized Midjourny to generate high-quality concept art and storyboarding at a fraction of the cost, without any negotiation. Consequently, we ended up handling only the production phase of the project, feeling somewhat disheartened as we saw preproduction tasks being automated. So what happened to those 4 Artist?
This trend continued with another project involving 3D animation, where a reputable agency suggested using Wonder Dynamics AI as its reduce costs by up to 70%. So no animators, no lighting and composting just some VFX prep.
Now we receive from most of the clients storyboarding and concept art using AI, leading us to rethink our need for 2D concept artists.
Hey thanks for sharing, those are totally valid points. I didn't mention it in the video, but my opinion for 2D artists like storyboard, concept, and matte painting is a bit different than my opinion with 3D artists. I think AI is great at those specific tasks, but I disagree when that idea makes a big leap in overtaking an entire 3D pipeline soon. Once you start getting into customized CGI territory, that's where it falls apart pretty quickly at replacing an entire pipeline. I'm sure over time some tasks will become more efficient (like making secondary animation, improving compression techniques, helping texture artists select base images more quickly, and helping modelers with base models to refine), but those optimizations don't go the full mile like it does when compared to 2D artists and copywriters, and AI may never get to that point for certain tasks. But yeah, it's interesting, and I suppose we'll see where it all goes.
@@cgforgeYou are right saying that less labor bumps the demand, but the curves (of offer and demand) don't necessarily overlap, I don't think the 4 artists mentioned in the comment would get jobs cause the demand have been multiplied by 4x, the same reasoning could apply in 3D, it's not necessary that the demand will follow the reduction in labor.
Yeah, @saedhmerakeb7089 I agree that the forces which move supply and move demand are not always the same. I was talking about the dynamics of supply when people decide to leave the industry. Less demand might exist for certain tasks (like concept art, story-boarding, etc). but if that causes productions to complete faster, then that may increase demand in general for productions, and that creates larger-than-usual demand for the less affected 3D artist jobs.
@@cgforgeYes, the question is ; is that larger-than usual demand offsets the decrease in labor (I don't have the answer) to me the answer is to encourage 3D artist to learn the new tools, Houdini btw have alot of AI application, btw do you use any AI (just curious about more "pro" people than me doing)
At the risk of sounding rude - maybe you should look for another kind of clients? I myself just had client who constantly kept coming back with some AI spit wanting me to make something similar. Finally I said politely that he probably should just use that AI spits he has and we parted ways. Not every client is worth keeping. People looking for quick and cheap anything will always be there. It took me long time to learn letting some clients go and say "no". For many reasons. Before it was "look what my son did in photoshop" or "my wife wants it red" and now its "look what AI spit I got". If somebody is stupid enough to trust AI to make storyboard that I can only pity the end result. But of course some will. People will always chase quick and easy money. Some will read a book, some will scroll tiktok videos. For those who choose tiktok you dont really need quality. AI spit will do. I prefer working for those book readers. Theres smaller number of them but also - number of people who can deliver what they are after isnt that big. More AI? Well, more mediocre jobs around.
I think it's true that AI will not replace 3D artists in the sense that it will duplicate their roles- the more likely scenario is that it will simply make their roles irrelevant. For example- SORA does not 'do' vfx- it simply obviates the need for them by implementing an understanding of things like fluid dynamics and soft body interactions at 'render' time- in SORA we see the potential that the whole production pipeline is compressed into a seamless process in which design, animation and rendering are simulataneously realized- in the same way that a Midjourney image comes into existence not as a linear sequence of sketch, final drawing, coloring ect but as a single emergent process where intial 'noise' is refined into a coherent and detailed result.
Your position here may be a bit like the blacksmith whose certainty that no machine could shod a horse the way he can blinds him to the oncoming future where horses are replaced by the automobile. AI will not 'take our jobs', it will simply obviate the need for those jobs to exist in the first place.
As an Artist I was certain that no machine could ever do what I do when I create an image- and that is still true- but Midjourney does not do what I do, it does something totally different- yet in the end the images it creates can in some senses compete with my own creations. The point is that 3D art is not normally an end in itself- it is usually a means to an end. If a cheaper, faster and thus more efficiant means to that end becomes available then it will be that technique that will prevail.
On a surface level, that all sounds correct, but it doesn't hold up in my mind once you get into the details. Clients want customized results. If they didn't, then they'd just go with stock footage / images. 3D artists are in the business of customization. I spoke a bit about an example in another comment that provides a specific scenario. I'll paste it here...
"...However, there will reach a point where some clients start getting picky, (most of them do), and that's where things can fall apart with the AI scenario. If, lets say, Geico wants their lizard dude to do burst through an exploding wall while he's dancing in mid-air and the fire is a little bit off, how is AI going to be able to go in there and help with those granular edits? If you re-prompt it, then perhaps the dance will be off or the lighting on the lizard doesn't match properly with the timing of the explosion, or there's a weird little bug that happened for a split second where the AI got confused and now it's affecting how it interacts with the next few frames, and for some reason the lizard dude's face changed in the next sequence and etc etc etc.... There's a million little deal-breaker details without a solution."
By comparison, just about everything has a solution when customization is the goal with CGI. So, if you're in the business of selling customization to clients, wouldn't you want to use the methodologies that provide you the most control? - Of course you would. You never want to be in a scenario of telling a client that it's not possible. It's not comparable to what's happening to 2D artists because the medium is completely different. It works differently, it has different goals, different reasons for people engaging with it... It's like comparing apples to tricycles.
I suppose time will tell if this is more like the blacksmith analogy or more like the flying car analogy.
@@cgforge What you say is true- I would go further and make the point that using text to define images is inherently flawed because if words could really define images in a precise way we would not need the images, the words alone would be enough. So the whole idea of 'Text to Image' or 'Text to video' is a bit ridiculous.
But in the real world where time is money and AI can do the job far cheaper than a human there will be a real temptation to go with what the AI spits out, even if this is not exactly the result you were looking for.
There are two possible scenarios- one is that the outputs of AI are so superficially pleasing and easy to create that the market will simply adapt to the limits of the technology and accept those limits as a commercial inevitability. In this sceanrio human artists will be seen in most cases as an unaffordable luxury.
The second scenario is the one you suggest, which is that over time the limits of the technology will become less acceptable as the novelty wears off and human artists remain relevant. It's not clear to me which of these will turn out to be correct.
There is a tendancy I notice among Midjourney users to 'retrofit' their expectations to the outputs their prompts deliver, insisting that what the machine made was precisely what they intended it to make all along, despite the fact that these systems have randomness built in to their algorithms by design. This kind of self deception may also be a factor that plays into the acceptability of AI Generated content especially if clients take a more 'hands on' approach and start to use the AI's directly.
Well, ya know, when it comes to clients wanting to save money instead of customizing, that's where I go back to the stock footage argument. If they wanted to save tons of money and that was their main priority, they'd just go spend 15 bucks at Envato and call it a day. What's interesting is that both the vfx / animation business and stock media business have been steadily rising together for many years now.
And that's a great point when it comes to Midjourney users insisting that the machine made everything perfectly the way they wanted. In the 3D biz, what the artist thinks is totally irrelevant though. The trick is to make the client feel like it's everything they ever wanted, and that's a totally different challenge.
Thanks for sharing your thoughts!
@@cgforge What seems to be happening on the 2d art side is that clients are getting increasingly 'hands on' and bypassing the Artist completely by using tools like midjourney themsleves ( Take a look at pramodnirwan's comment below re how their client used Midjourney to do their own concept work) I can't see this happening on the 3D side anytime soon, but the rate of improvement in AI at present is genuinely shocking, so who really knows?
What's so unexpected about the whole generative AI thing is that it's not really replacing human skills with machine equivelent skills, it's eliminating the need for those skills entirely. Your arguements make the assumption that clients will continue to feel the need to interpose a 3D Artist between themselves and the final output, and as long as that remains true I think 3D artists are safe. But if the technology matures to the point where a client feels that they themselves can create exactly what they want ( or at least convince themselves that they have done so) then all bets are off.
The real problem with AI in general is not that it is really as competant or capable as it's made out to be- but the results it can deliver create an illusion of competance that can be very compelling to people whose ability to discern good or bad aesthetics may not be that well developed. The result being that in place of genuinely skilled and talented creators we get dillettantes armed with AI who fondly imagine that the medicore derivative outputs they can create are in every way as good as those they once obtained from paid proffessionals.
@@paulhiggins5165so true
When people are talking about AI and taking jobs, they're not talking about tools that are assisted with AI that make working faster, they're talking about generative AI. Right now, its focused on producing images, but eventually that will include creating 3D models, animations, code, and much more. Then it will be replacing artists and then see if you feel the same then.
Your argument about flying cars vs what computers what will be capable of... well, not a very good comparison. Flying cars have real energy problems, but computers have no such limitations on producing content.
And as far as the AI producing something new... we all know that Hollywood is just a recycling machine, studios are more than happy to regurgitate the same movies over and over, its rare when something original comes out.
AI is going to dissuade a lot of future artists from going to school or entering the field because a lot of it will be able to be done without them.
My point with flying cars is that many people didn't understand the limitations of car technology when it was first invented. Nobody respected the fact that reality has constraints on how things can work - not imagination. AI is already capable of incredible things, but we don't know what's possible with it. It could be near it's max potential right now. It could have a long way to go. But it's a big assumption to think that it's just going to replace everything, and many people stop thinking at that assumption.
How do you know what eventually will happen? :D "It will do this, and then that, and then even more!". Its just repeating same slogans but - shockingly - NOBODY KNOWS THE FUTURE. Same people who say "we have never seen some technology like that!" are claiming that they know what will happen in the future :D Based on what?
@@tlilmiztlibased on the fact that tech layoffs are already rampant due to AI. It's already happening, we don't need to speculate
I agree. AI is going to displace a lot of workers in the same way automation replaced auto workers, cranes replaced dock workers etc.
It's not that AI will replace all human works it'll mean that one person can replace 10 people, losing 9 jobs.
This video is helpful. 👍 I also love how you added the "They took my job!" gag from South Park. 😄
Nice positive take, thanks for that. I disagree with some of the conclusions though. Yes, demand for digital visuals will keep increasing, and will accelerate, but let's be realistic. Look at some of those things sora is already putting out. If it can animatic a still image of a wave and within hours get to same level of quality that would take a couple of artists days or weeks... then nobody is going to hire those artists, they will hire the 1 artists to do the AI prompts and then comp out the mistakes of the AI generated video.
So yeah, there will be more demand for certain types of artists (prompters and compers) but less demand for things like FX in this case. This will also be industry specific, since i don't think that will impact film as much, but in advertising this is definitely going to require less FX. I recently worked on something where we literally had to recreate a print add they made, but make it move, look had to be 1-1, just with 3d sims and stuff. Sora would do that a LOT cheaper.
So while I agree it will create a lot of jobs overall, it will definitely also replace a lot of jobs. Have sora generate a flythrough of a environment, track the cam and get a mesh from it, build the whole thing in comp. So yeah, more work for compers, less work for 3D artists.
Hey Tim, thanks for sharing your thoughts. You made a really good point about shifting demand towards prompter artists that's being taken away from 3D artists, and to a limited extent, I would agree in certain scenarios. The main catch there, however, is that the client needs to be cool with having less control. So, instead of a client having, lets say, 75% control over the specific end product via the normal 3D workflows, they need to be cool with, lets say, 50% control.
For some, that's no problem. They're looking for a quick bit, and AI is a win for them in that scenario. However, there will reach a point where some clients start getting picky, (most of them do), and that's where things can fall apart with the AI scenario. If, lets say, Geico wants their lizard dude to do burst through an exploding wall while he's dancing in mid-air and the fire is a little bit off, how is AI going to be able to go in there and help with those granular edits? If you re-prompt it, then perhaps the dance will be off or the lighting on the lizard doesn't match properly with the timing of the explosion. There's a million little deal-breaker details without a solution.
If, in that scenario, production hits a wall with AI, then now they're going to be scrambling with how the heck they're going to fix up that fire which sends them running back to simming an explosion via Houdini and potentially starting from scratch again. Over time, I think clients, too, will get stung by the lack of granular control, and that sting could lead them straight back to doing things the "Ol' fashioned way" and realizing that AI isn't designed to be used in that capacity. Over time, studios and clients will discover those risks, and a lot of them will adjust back in the opposite direction. Over time, AI will become more sophisticated, but even it can provide granular edits like that, it will be in an environment where many professionals have been burned to some degree by AI and probably skeptical to risk it again. I suppose time will tell!
@@cgforge If we look at how this works with stuff like stable diffusion:
Things like control-net lets you control posing of a character. So you would input a animated character (yeah that would still need mo-cap or hand animation) and that would drive the AI model. You can make slight light tweaks the same way as well, this type of stuff already exists. You have plugins for various 3d software that lets you “render” with stable diffusion where it takes a basic input scene and lights and renders it with AI.
Of course there will be exceptions to where this stuff can be used, and there will still be 3d artists necessary, so your point in that the people that do stick around will be probably able to be paid reasonably well. But a lot of the traditional 3d jobs like fx artists, would disappear. Jobs where you control the creative direction do or help someone achieve the creative direction by knowing how to manipulate these AI models would have a lot of staying power though.
And yeah for the very specific clients that want to pixel fuck, traditional way of doing things would still need available, but given how it would be 10 times more expensive than the AI alternative, that would only be for the very big clients.
Incredibly helpful video, thank you so much for taking the time to make this. Can I ask, on the instancing test, how did you get the "instancefile" attribute for redshift proxies method working in Solaris? I haven't been able to figure that out yet!
Hey @kylehamm8928 , all I did for that was bring in the points from sops that had the instancefile attribute, and that automatically got picked up by RS at render time to grab the RS Proxies from disk. My guess is that perhaps there's an issue with your pathing if it's not showing up.
When assigning the attribute to the points, I used an attribute randomize node in SOPs set to "Custom Discrete" and "String." Then, my string values were set as a relative reference of the paths described on the RS Proxy Output nodes. The resulting path was then absolute (meaning that there wasn't any $HIP or $JOB variables in there) and that did the trick.
This is pretty good take on the topic. I appreciate that you're not getting swept up in the "drama" and trying to look at this objectively.
With that said, I think there's a lot of us that aren't worried about what AI can do today, but rather what it'll be able to do in the future. While I get your argument that you can't truly know how the future will play out (and this is also why I'm continuing to treat my work in VFX the same I have for many years now), when you start looking into how this technology works you kind of realize that some of the "worst case scenarios" with AI not augmenting, but fully replacing workers not just in VFX but in general, is pretty near inevitable.
So, I agree with you that as of right now, there's no point in panicking. We likely have at least a few more years before we really need to worry about this. I just encourage people to look into and educate yourself in AI, and more specifically how all the current AI tools will ultimately very likely lead to AGI. I don't view this as a challenge for VFX though, but rather for society as a whole.
For now I decided the best thing I can do for myself, is to stay educated on the topic and utilize the new tools in productive ways, and keep working and creating.
ever since you mentioned about this video on your live stream, i am waiting since morning for this video
Thanks for checking it out @OhioNPC911 I definitely ruffled some feathers with this one. haha
@@cgforge yes, really enjoyed the video
And I realise one point about marketing , nowadays its a weapon to hyper anything without knowing its true potential or limitations
You’re the man Tyler! Really good points here
Nice video Tyler! I also believe that with AI coming around, there's a great potential for the development of new tools within 3D software that can substantially enhance workflow, like a much more efficient way of rendering and denoising, and some new tools to help in animation, modeling, and this kind of stuff.
Exactly, you beat me to it.. especially the rendering part.. it's going to help reduce rendering time while maintaining same quality or even improving the ray tracing... likewise easy Roto and Keying.. it's not just about the language models and image gens
I agree! For animators it'll be awesome because it can create realistic secondary motion. Faster iterations without the tediousness = more iterations in general = higher quality across the board.
In Music and 2d illustration (non generative) AI was helpfull aswell. It's generative AI that is displacing artists. And generative AI for 3D doesn't exist right now, but with 3D Websites selling User content to AI companies, will most likely arrive soon..
It's important to realise that AI is scary for artists because we can literally see the performance increase. We can see midjourney improve in quality many times over in a couple of years. But it's important to realise it is making similar strides largely unnoticed in other domains. For example a risk manager at an asset management company has no idea if AI will replace their job. At least we can see how quickly this technology is developing, see the hurdles and at least get a bit of a head start.
I really like your take on this, Tyler. I couldn't agree more with everything you say. I myself have let myself be down the rabbit hole of anguish and misery, but then something clicked and I realized that the current AI tools will definitely not replace us 3D artists. I appreciate you taking the time to promote this kind of logic and positive thinking in an industry that's at its most negative these days.
It was refreshing to listen to your point of view. A lot of bad omens speech has reached my ears, it was kind of affecting me. thank you.
These ai generators don't have any kind of precise control needed for serious work, unless they can read directly from our mind. Houdini would probably implement a lot of ai nodes(tools) in the future versions, so our 3d skills and knowledge won't become completely irrelevant over night. It's like these people in the comments don't have basic understanding how 3d and cg works
I agree with you 100%.
Thank you for this video.
Like your positivity! My main concern is proper compensation for all the artists whose work is used to train the models. Hoping this will all get worked out to the artist's satisfaction ...
Indeed... artists have always been in a pinch when it comes to negotiating fair compensation.
Thanks Tyler
Thanks Tyler , sharing for this … I always enjoyed your teachings…
Biggest issue right now for any artist is that “ what is the purpose of learning so called traditional way of learning “ because some machine can do within a second the image which the artist actually thought in mind and was prepared to go the journey of achieving the result by manually creating it
Art work / musical /writing or any creative technical work is a journey and that is consistent for any human
A human should have some purpose to live and go beyond ..It’s not just job satisfaction.. it’s a journey that is what is important.
We human can not compete with machine power .. never going to outperform them in terms of speed and calculation. But this is where the big question as this is reality and artwork , creative works takes time to
prepare .
Also the profit making industry will always come pounding on you take the last freedom you have ..
In reality many people are learning AI skills just to save job without having any interest in it . For some it is just another avenue of making money
It’s another rat race for very vast section of people on earth .
In My opinion AI should only be used where it is really needed like .Robotics , medical , life threatening job etc
AI should be nonprofit entity ..but in reality it is going exactly opposite and big corporations are racing for that too
As VFX is concern where many tools from AI will be surely great to use and that’s completely ok
Concern is exploitation will increase more rapidly…
I can only hope that it does not go that direction
Awesome Video! I completely agree with your view on AI.
Thanks for watching @CGINerd
I half agree on your points Tyler. The things that you are correct in my opinion are that yes as a tool ai can and will speed up workflows, enable smaller companies to work faster and be more efficient. But... The point that most people are bitter about is how AI is trained and we get to the results we do. There is no way to prevent ai to train on your artwork, no way to protect a unique style that you are known for. And protect is probably the wrong word for it but what i mean is people some times go to specific artist because they want their unique style of doing things and if an AI can generate this with a prompt, without the original artist to be compensated at all then in my opinion that is just theft. Because like you said, AI does not invent new things its a very good kit bash machine. By that definition it uses your work directly into making other things without consent. I recently found out that my 3D renders have been used in training AI model and I am not happy about it because no one aske for permission, no one offered any sort of compensation etc. The biggest grudge people have with AI is the generative part of it. If you can come up with a tool that can speed up GI calculations, remove the need for UV work, or any of these thigns then no one would be mad about it , in fact everyone will praise it! But the current implementation is not very ethical and this is the biggest problem people have with AI in my opinion. Its not even the fact that it can generate images and videos from nothing its the way it is structured to scrape people's work in order to be able to work. :)
Pretty much what I have been saying, and my point is MANY people really don't understand what AI does and just gets impressed by the advancement and the ( selected ) reels. They don't understand the basic premise of 'training' the AI to do something, they don't get that part, they see a beautiful video , they assume EVERYTHING is possible ( and will be CHEAP ) .
Same with some clients now, they start injecting AI goals and ideas , like is something you can just 'implement' out of the box in Pipeline, animate without constrains, or create this or that. Is either AI or Unreal, two things that people have mixing ideas . Many clients keep talking about Unreal as if it was some magic tool, and think you can throw everything at it. And Unreal is cool, but NO, is not like that, has limitations specially when you start rising the bar. Ai is the same, works for the ideas and sketches, but you will need BIG pockets to enter and 'train it ', and specialize it, to do what you want, and hope that if you reach that goal , is what you will need for a long time! Because you will need something else right away, Is like Building a whole automatic physical Factory for custom 1900's cars,.. for those pesky 30 cars you will need for a movie, great! next project is about Life in Space... Good luck finding purpose for that factory.
Somehow reminds me of the CGI Ruse of the early 2000's, many producers had no clue how to get into it but they wanted a piece of it, and producers kept thinking you could just put an actor in a empty green room, and table..( to avoid production costs) and everything would look 'realistic' around him if done on PC. Big news, at the time was way cheaper, and clean to dress the actual set , than to recreate ultra realistic lighting and interact with objects. With time they came to realize it was not 'magic'.
Those are really great analogy's. But you only think that because AI has not replaced 3d tools ... yet . And at this rate it will be able to replace a lot of background prop creation and mid ground work. Which will mean we will only have senior level animators, character modelers, riggers, programmers and so on. So there WILL be more job losses in the near future and competition for those jobs will sky rocket as it has in the last 10 years. But i think in the long term you will be correct and it will just mean that we will be able to create more content and the demand for content will increase as will demand for people who are skilled at creating that content quickly. I think quality will drop. Much in the same way that animation quality has dropped since the 90's for kids cartoons broadly speaking I do not want to generalize. I also do not think AI will be good for wage growth. The trend since the 60's generally is lower wages and AI is going to accelerate that trend and continue accelerating without massive changes to the tax system.
Really enjoyed this. Thank you!
well said and good points were made. just like softimage didn't upend animation, AI will spark productivity - BUT will make certain steps obsolete. My main concern - and that goes waaay beyond the use of AI in entertainment - will be how accessible AI/ML is going to be and if we're not just looking at yet another monopoly where only big money allows for big tech and any effort just goes towards the perfect subscription handcuffs rather than furthering society.
I agree with most of you said - there is going to be a period though where producers will want to use AI early on to save their beloved bucks and there will a period of frantic hysteria in an industry that's already fully saturated with hysteria =)
Valid arguments. There’s obviously pluses with this tech. The thing is the ratio of those pluses vs minuses with AI. Generally AI is just another tool. How it’ll be mostly used is one of the questions. The rapid improvement of AI doesn’t seem to be slowing down. So, I think you’re right that it will be beneficial for midsize/small businesses. Although I do think AI will bring more problems, than solutions. I’ve seen a lot of people loosing their jobs in another industry as a “optimisation strategy”. It’s not just AI, but it’s the AI that made it possible for such large scale. Like in simple example, a company can have less people do more work.
Overall it’ll be interesting to see how it’ll go forward.
Yeah, most definitely - I agree that there'll be a lot of pros/cons that end up happening with it. I didn't mention it as much in the video, but it would be interesting to know how many of the current layoffs are due to AI and how many layoffs are due to a combination of factors like outsourcing, tighter lending conditions, and an overall decline in viewership due to streaming wars + less movie ticket sales. My guess is that many employers are saying that, "AI is the reason you're getting fired," but in reality, there's probably a lot more to it than that.
I'm not "slyly tiptoeing" around anything. My comments here aren't contradictory to the opinions made in this video, and I'm happy to have an honest dialogue with those interested in having one.
Need more episodes
Hmm... looks like I need to find a more controversial topic first. "Unions" perhaps? I'm sure something about election season will also do the trick. haha
People spend years and years learn houdini if there is no future people learn for future
as someone who would like you to be right about this, this is ridiculous. ai is going to be more intelligent compared to us, than we are compared to dogs. that's not an exaggeration.
"30 texture artists"
ai doesnt need a single texture artist 🤷🏽♂️ Until then, like you said, ai will make things "more efficient" which means *less artists needed*. Did you think studios will keep paying the same number of artists, plus ai software?
why don't you try these foolish arguments on the guys who draw 2d animation, stop motion or practical models... oh wait you can't, nobody does those anymore
it only took 1yr to go from worthless ai video to viable video. Imagine just 1 more year.
Katzenburg ( dreamworks ) was quoted last year saying there's only 3yrs before ai full features.
half this video could be simply summarized "think positive!!" 🙄 That only helps in situations where you have 'some' control.
"reality has limits"
yes and there's a real limit on how much content the market can support. If we are to keep working alongside ai, there'd be 10x more movies & tv shows.
Well, I still think we have a bit more time before those AI robo overlords turn us into dogs. And, similarly, there is a very, very long way to go for AI to overtake all digital media. My message isn't just "think positive." My message is that there are strong counter-arguments to the doom/gloom argument, and not enough people are thinking critically before dissuading themselves or others. Promoting "positive" conversation is to say that we ought to be taking a more detailed look at the situation rather than just typing out our feelings of doom for everyone else to read.
@@cgforge
And it was a "long long way to go" before it could do 'moving' images. Turns out 'long long' was just one year. It's already good enough to start replacing everyone who does stock photography / video. Do you think its worthwhile to study 3D, if you can only do it for 3-4 years?
And sorry, no those aren't "strong" counter-arguments. We already know that "scarcity increases value" but that's not true when demand has dropped, and that is what is happening. 2D & stop motion animators are not paid better, just b/c there's so few of them now.
I think the main area where we disagree here is on how similar 2D artwork is to 3D. In my opinion, it's not even an "apples to oranges" comparison... it's more of an "apples to tricycles" comparison. The two mediums are much more different than people think they are. A client is going to hire 3D artists for completely different reasons than for hiring a 2D artist. Generating ideas is typically the goal for 2D artists whereas customization and fine-tuned control is the goal for 3D artists.
Let me provide a practical example: Company XYZ wants to showcase their new product at a trade show. Everything has already been designed. Prototypes have been made, and the concept portion of the pipeline has been long established. Now, it's the job of the 3D artist to make Company XYZ's dream come true by making a cool reveal animation as the CEO is talking on stage. We have a product that nobody has ever seen before. It needs to look EXACTLY as it is in real life. And, Company XYZ wants to get realllly picky about everything because the CEO has been dreaming of this moment for the past 5 years.
You really think AI is going to be way to go here? How is the 3D artist's job similar to the 2D artists job? How are you going to train it to create something that's never existed before? What happens when AI gets the dimensions of the product off jusssst a little bit? What tools does the 3D artist have to fix these issues and made sure that each shot has perfect consistency? Again, customization and fine-tuned control is the name of the game for 3D artists, and their work is incomparable to 2D artists.
Even if AI continues to make incredible leaps in progress, it's not going to impact 3D artists in the same way as 2D artists because it's a whole different world of considerations.
@@cgforge
AI doesn't need to do 3D. 🤷♂
Have you not seen the woman walking down the street at night in Tokyo. No Modeler, no Rigger, no Animator no Texture artist no Cloth FX artist no Lookdev artist no Lighter no Compositor. That's 8 artists out of work right there. Just a VFX supe and director typing prompts.
Just how you're seeing 'more' work - or even just no loss in amount of work - in all that, baffles me.
"made sure that each shot has perfect consistency"
AI is already doing better medical diagnosis than human doctors are. Computers have always been more consistent than humans.
" product at a trade show"?
3D graphics have been notoriously off the actual product, by at least as much as AI would be. That even happens with photographing the actual product. And that would probably be far easier to re-do using AI.
You sound exactly like guys a year ago saying "HOW is it going to do video?"
Do you rly think it's better to think "there's no way that new tech will take my job, just like it did for all the trad animators, stop-motion animators, model builders and FX people"-?
But yea I get it you wanna sell 3D training as long as you can
Interesting look at the situation. I am more of a hobbyist and not one in the industry. So, I believe that I can make a neutral observation here. I do agree with you on many points. I do believe that AI will only make things better in the future of 3D. I also believe that time should also be considered. AI is unable to do a lot of things I am looking for. For example on one AI program I tried, it stated that the model I made was holding a pink object in her right hand. She was wearing a red bracelet on her right wrist. A friend of mine is a manager of a company that does a lot of engineering work. He stated that AI would eliminate 3 individuals from his company and make all the rest of the individuals (around 20) much better at their jobs. So, AI would heavily benefit most of his company, but 3 individuals would lose their job and potentially have to learn a new skill. AI is in its early stages of development. If it does as you say, and open up a lot of jobs, then where are these jobs going to find workers ? I believe that it will take workers that were laid off in other industries and make them learn a whole new skill. This may not be a bad thing. Technology is always changing. And we as a people sometimes have to change with it. AI will do several good things for this world. Overall, It will be a lot better than its few drawbacks.
You can't look at the AI issue in isolation. You have to look at it in the context of a post-2020 landscape. The Hollywood "Multi-billion dollar profit-machine of a movie" model has collapsed. Franchise film IP has been mismanaged (think SW or Marvel) and steered into the rocks. Badly timed strikes have compounded the rot.
Therefore upcoming budgets will be smaller and CG & FX will be dialed back.
The Ads industry seems to be contracting also.
I agree with you in general. Much bigger issue has to do with market forces in general rather than AI.
Hollywood has definitely been over corporatized and resulted in movies that many people don't really care about as much as they used to. It also wasn't terribly long ago that the Netflix revolution changed the business model entirely for movies, and that's resulted in everyone trying to undercut each other and creating a more chaotic market that's less fun for viewers to sort through.
On the plus side, I do think there's a lot of upward potential in episodic TV that uses lots of CG. Game of thrones was obviously a tragedy with how it ended, but it did give a glimpse of what could be possible if you had long format stories that didn't get ruined working with CGI.
I suppose only time will tell where the new opportunities will sprout up, but in the meantime, it is definitely resulted in less work for many people in the industry
It's possible for AI to create things that never existed, because just like human imagination, it's still based on existing things (for example : xenomorph aliens. Pure creativity out of nowhere is a myth)
I saw someone else on RUclips saying the more AI become good, the less it's creative and it's true ! (Check out the video "Ai 'ART' will get WORSE not better" by Sam Hamper)
I noticed a fun example about this : stylish lingerie xD Last year Stable Diffusion would create amazing lingerie that I've never seen in my life, with lots of mistakes but still beautiful and creative; but now that it becomes more accurate, it's starting to look just as boring as real life lol
I must admit, I was terrified of AI in its beginning. Until I took graphic design class with a teacher recommendation from my animation professor. Glad he did, we learned how to use photoshop and actually use it which was plenty of fun. We even got to mess around with adobes A.I tool for 4 of our projects for spruce ups on them which was fun to learn. She was all about stepping up in using the current tools in our disposal. Which made me feel better with my doubts. I've been using A.I like chat gpt and others on my phone ever since to help me get creative ideas for storytelling, painting ideas when I'm stuck in jams. ever since. i find it a godsend. when stuck in a jam to boost up ideas. I even use it sometime to help out with my coding in my gamedev learning journey once in awhile. I've been noticing it has helped me tons on speeding up on the learning aspect.
dude, look at how many industries (jobs) disappeared overtime because of tech advancement (i.e : graphic design in the 80s-90s, and is just one of many many many example), most of your arguments are easily countered. I've been 25 years in the vfx industry btwn. I could go on and argue with counter arguments but the toxic positivity, delusion is unreal and pointless to argue. We ll just have to see in a few years. The main upshot for us is, Ai and copywrites, that's it. nothing else. Everything else, artists are on the losing end. And deleting comments that don't support your narrative is not an open realistic conversation but exerting control over people that don't agree with you.
I didn't say that I was going to delete comments that disagree with me. I said that disrespectful, trolling comments aimed at spreading negativity and/or that bully others will be deleted. So far, the comments on RUclips have been great for the most part, and I haven't deleted anything yet.
So, if you're interested in providing a counter-argument, then what would it be? If you've been in the industry for 25 years, then I'm sure you must have something insightful to add the conversation that goes beyond mudslinging against my opinion.
I have told illustrators for years, there is something coming. Seems like alot of people only realize it, if the real displacement is there. I don't see a valid point, why AI shouldnt be able to do 3d models?
A concept artist doing a concept that is translated to 3d via ai can even be copyrighted.
And why they shouldnt be able to incorporate into something like sora?
As 3d replaced most of stopmotion, ai will replace most of 3d.
The problem is that comparing 3D artists to 2D artists is an apples to oranges situation once you get into the details.
AI will not eliminate the need for artists. But it will allow one artist to do the job of a 100. People should be allowed to be concerned about their livelihood, dude.
I disagree with the idea that one artist will do the job of a 100. Let me give another example - is AI going to replace the job of professional photographers? Well, no, not entirely. Why? Because professional photographers can't just type in a prompt, "Give me photos of my client's wedding" or "Give me photos of my client's dog running in a field." They get hired for customization. It needs to be that exact wedding and that exact dog running through a field. It can't just be any dog or wedding. The same idea goes for CGI artists. AI is impressive technology, but I think many people misunderstand why people pay CGI artists in the first place and why AI is not going to replace those services.
Finally somebody sane :) Yeah the AI hysteria is really getting old. Its not even AI for crying out loud... And I feel like people who talk about future of it didnt really tried to use it for real. If they did - they would knew how limited it all is, how inaccurate. The more specific is your project, the more specific need you have or idea - the less this pseudo AI will help you make it. People talk about how it will soon (its always "soon" or "in 5 years, you'll see" :D ) it will spit perfect 3D models :D Dude, it cant even make good retopo for crying out loud. Eh... Thank you for being voice of reason in all that ocean of mindless panic and baseless predictions.
AI is not a job replacement; it's an industry replacement. It's akin to telling car manufacturers that people can now create their own cars using just text.
That may be true for the stock footage industry, but it's a pretty big leap to say that the multi-billion dollar film / tv industry is just going to keel over real quick because of some flashy tech demos.
I hope you're right! But for people who are working in the big comrporations and studios are already losing their jobs on daily bases in their hundreds to A.I.
Repititive jobs which were perfect for those not too tech savy or for example rotoscoper jobs are pretty much extinct!
Perhaps this video, I assume is all about not to make already dumbstruck emploees not to fret too much and find jobs in other sectors and areas so to keep the investors stil interested to invest in VFX/CG industries so you and I can keep our jobs. In this case I salute this video.
thanks.
Thanks for watching @No_Plastic. For those losing their jobs at larger studios right now, I think that it also has to do with a greater degree of outsourcing, liquidity getting tighter in general (which makes it harder for projects to get funding), and streaming wars that have led to lower quality productions. A studio might say that the reason someone is fired is because of AI, but it may not mean that's the only and/or main reason for it.
The point about flying cars ehh, it's mostly about regulation than anything else. other than that I agree with everything you say in this video. The outrage is completely out control.
😎😎😎😎
what this guy said. ⬆⬆⬆⬆⬆⬆⬆
New vray will use ai generative tool to create textures inside the software. Good luck to those 30 artists :)
That's a good point, but my counter-argument to that is why didn't texture artists go to the way-side when substance painter came onto the scene? Substance was a HUGE time-saver over the previous workflows of UVing everything perfectly and relying on photoshop for everything. And yet, demand didn't dry up for texture artists when that technology came around. Faster iterations = more time for feedback / changes or a production getting completed more quickly. That increases demand for movies / tv shows because either the quality goes up or the speed goes up. That means more projects get on the docket for studios, and that means more demand for texture artist jobs.
I recently embarked on my 3d journet in blender and im addidcted. Really good insights here. Thank you sir!
seriously, this whole thing reminds me of when photoshop was new. first, you had people saying “digital art isn’t real art, you just push a button!” (the same thing was said about photography in general in the 1800s), then in about a decade, no one wanted to hire film-only photographers… but while you might have heard a photographer griping about needing to learn new stuff, you didn’t hear “these new-fangled digital photographers took my jerb!”
honestly, when you see all the ai haters, the comments are full of self admitted communists who completely lack any economic knowledge whatsoever. it’s just jealousy fueled hatred. especially since the same exact people couldn’t shut up about wanting ai servants ten years ago, but now that it might replace THEM instead of the “low skilled workers who are doing ‘jobs no one wants’” (i can’t stand that phrase) so it’s the most horrible thing ever, and “it violates their right!” (it doesn’t.)
seriously tho, if ai takes your job as an artist, your job probably wasn’t as important or as integral to that company as you thought it was. and considering how many people in the “ai is evil and it took my jerb!” camp ARE spiteful communists, i feel like a LOT of crucial information is being left out of their stories. (they never have any real proof, just a lot of claims and anger.)
either way, luddite movements never succeed. the people will always side with what is more useful to the whole (something the communists claim they want, and yet they rage when they find out maybe they weren’t as irreplaceable an individual as they thought they were… which they should have expected if they actually understood their own ideology) of society instead of what is preferable to relatively small group. ai art tools are going to continue to progress and become more powerful and intelligent whether or not people fight against them.
He is definitely right, nowadays as an example the demand to contemporary paintings is like never seen before! Everyone want to buy contemporary paintings, literally everyone! Probably you all also have a few on your walls. Jokes aside, where is the AI driven UV tool, retopo tool, rig build up tool, subpixel dynamic render calculation tools... ? Where are they? :) I believe that most of "AI" development is aiming to deliver the final product, and they do not want to help in development (not all but some of them) So yea, the good moral is an important factor, lets all keep up, but lets not start to draw a wolf as it would be a cute rabbit.
So lets talk a bit about imaginary vs real. The reality of it is that you are in the business of selling courses to future or current 3d artists. As you mentioned some are off put by this ai talk and might reconsider a career as a 3d artist. That means that you are biased toward downplaying this ai thing to keep that revenue coming from selling courses.
Having said that the reality of these ai tools is fluff, all talk and no show, sci-fi, etc. and im speaking from experience and from seeing what others do with it. At best it can do some half decent concept art after numerous requests.
The reality of it is that they dont even enhance or enrich the life of an artist because we dont have tools tailored for the artist, most of them try to circumvent the artist from the process. No ai UV mapping, no ai material generation, like why is it so hard to train ai to extract roughness map or normalmap from a photo!?
The reality of it is in the other comments saying that the clients started using ai at least partially instead of hiring artists.
I am a spare time solo game dev that uses photoshops ai to remove defects from photos (because thats all its good for). i will never pay for something like custom game music because i have no budget for that but if i can get a tool that can generate something half decent for almost nothing... Thats the reality of it
Hey @teodorstaicu38 , yes, as you said, I'm in the business of selling courses. That doesn't disqualify my opinion or the points I made though. Even if I wasn't an educator, my opinion would still be the same. AI does bring up some interesting thoughts on how the CGI business works in general. As an example, "Why do people hire CGI artists?" I think the answer to that question is misunderstood by many people. In the next video, I plan on taking a deeper dive into it. Similarly to photographers, I don't think AI is going to replace them. Why? Because, just like photographers, CGI artists are in the business of customization. It can't just be a "wedding ceremony." It needs to be that person's wedding ceremony for photographers. Similarly, it can't just be any ol' racoon in Guardians of the Galaxy, it needs to be "Rocket Racoon." And he can't look different from one shot to the next. AI is good at generalizing ideas, but it's not good at providing specific ideas on a continual basis due to how the technology works. So, in that way, I don't think VFX artists need to be freaking out.
Man I really love your brutally honest and logical breakdown approach to everything
And IPs. IP ownership and NDAs are an issue with Ai created content. So using tools to help (denoising, mattes, roto, etc) is not the same as creating a production and people who don’t see that are very foolish.
I have a personal gruge against people who talking Ai taking jobs especially the ones who aint from the industry. I promised myself when i will be in the industry , would personally call them out and prove them wrong. Its so easy these days to openly say whatever you want , and they think its quite easy in the entertainment industry, As if its just a click away . They dont fucking realise that CGI and VFX is probably one of the most Complex Forms in terms of storytelling.
Especially Reddit and Facebook. Folks stay away from those! I repeat , Stay away! Saying it from the same reasons CG Force said , they will pour their own misery into you.
Yep, I agree - and those spaces have a lot of bullying going on that I think is worth pushing back against.
AI will take away human jobs, but not every human. The impact of AI will be to adjust the division of labor among humans, and those without suitable skills will indeed be eliminated in the re-matching process. I completely agree with CG Forge’s view because this is the truth, and there will also be another part of the truth. 3D does have more difficulties for AI to overcome than 2D, because 2D does not need to understand time logic, but all this is just a matter of the speed of technological progress. You should know that a year ago, AI videos were just of very laughable quality. Since 2D artists will be replaced by AI, there is no absolute safety for 3D artists. When AI can draw realistic 2D paintings, many 2D artists even looked at AI with a mocking attitude because AI couldn’t even draw fingers well, and this bug was fixed only within a few months. But having said that, strictly speaking, it cannot be said that AI has replaced these 2D artists. Perhaps these 2D artists are still engaged in 2D work, just using AI as a new tool, but inevitably their direct income may decrease because the disappearance of technical barriers makes the market more competitive. Some people say that the government will not allow AI to take away human jobs because they need job positions to create tax revenue. In fact, this statement is not valid. Just like after the industrial revolution, the government didn't care about workers who were laid off due to the popularization of machines, because the increased production capacity of factories created more profits and paid more taxes. Overall, I think AI will definitely have a big impact on the 3D job market, but AI will not make everyone unemployed, because there always needs to be someone doing debugging and bridging work, and there is still some distance from establishing a workflow centered on AI. People born in the era of technological change are destined to experience pain, just like every era of technological innovation waves.
Thank you for sharing your thoughts @ikeruzo
2D Artist here I'm not scared of AI. I don't think it will be original or creative. The thing is you need to train it on data that already exists. If it doesn't know a dragon it won't make a dragon. If you train it on cave paintings it will only make cave paintings. AI "Art" is all doomsday hype. I'm just going to create things instead of worry.
Thanks for sharing! I also tell folks that doing concept / 2D art isn't just a product. It's also a service. It's providing an experience to the client where you, the concept artist, are trying to capture the ideas / imagination that the client has in mind. That "service" of working with them is a people business, and that side of the equation doesn't change with AI. As an example, perhaps instead of 20 iterations, now you can do 75 within the same amount of time. That allows the client to have fun with more options, but that doesn't mean that they'll hire you less. It could just mean that they spend the same amount of time picking through more options. Does that align with your experiences?
AI is a valuable tool that is expected to boost productivity and support the growth of other artists. There is currently no data to support the claim that AI has replaced any artists' jobs. Do you know why? Because it hasn't displaced anyone's job. Utilize AI as a practical tool for enhancing and producing.
Completely agree... Now there is too much of tedious steps to create CG. Those things if AI can get rid and allow people to create more and worry less about technicalities that will enabke more people to realise their ideas.
Well said sir. I believe the doom and gloom perspective is being fueled by two camps: artists that are relatively new to 3D/video/design (and haven’t had much experience in Production, or comprehend how important iterating is), and jaded artists who are scared for a variety of reasons. It’s cool what popular AI tools “can” do, but what they “can’t” do prevents them from ever being used in a Production pipeline…at least not as a final deliverable. What popular AI tools do well: story boards, parts of set extensions, brain storming, etc. They speed up the grunt work, and give you more time to spend on the thoughtful work. Professional machine learning has been used for decades. Ask anyone who has ever worked at ILM. Once a popular AI tool is able to preform a PERFECT Roto mask, junior artist won’t have to spend the first year of their career Roto-ing. It won’t be that a young artist can’t get into the industry, it means they won’t have to wait as long to start getting more fulfilling assignments. We still don’t know what the legal implications will be for training content. we say that the stock photo/video industry is dead. i predict some of these companies are going to start hiring creatives to build their own proprietary libraries of photos and videos to be used for training. did auto correct put copy editors out of a job? No.
Late 90s when I started in TV production only a hand full of people had access to systems like Quantel or Flame. Let alone worldwide. Now everybody is going nuts about AI taking all these jobs. These jobs didn't even exist 20 years ago.
@@misanthrope_01 Nuke exists because Apple pulled plug on a software called 'Shake" back then out of nowhere. How old are you again, kid?
@@misanthrope_01 You can't throw a rock without hitting a 'designer' nowadays. And back then it felt like 500 people worldwide were making a living with media design. And now everybody is crying because of AI like the end of the world is imminent. Do you get the point now?
I remember conversations about digital photography in the late 90's. The argument was that it could never replace analog photography because it was incapable of achieving rich black and white values. I remember conversations about multi-touch interfaces when Apple released the first iPhone. The argument was that people would never be comfortable typing on a screen without tactile feedback. I remember conversations about Bitcoin back in 2014. People were comparing it to the Dutch tulip mania, ridiculing anyone who was willing to pay 150$ for a single Bitcoin...
Getting upset or angry at a technological revolution has never worked, you either adapt with it or you perish. The hard fact is that, the use of computer graphics, in non-interactive entertainment is coming to an end. Instead of fighting this, 3D artists should welcome the advent of emerging technologies like spatial computing. The transition to a more immersive, 3D based interaction with digital information will bring with it a plethora of new opportunities for skilled 3D artists.
Thanks for sharing your thoughts @stinosan. For 2D artists, I think the analogy of digital vs analogue photography is a good one. The digital camera took on the same exact product and role as what the analogue camera did. Looking back into the past, every highly successful technology always has a group of nay-sayers that were proven wrong. However, there are also many technologies that we forget about because the nay-sayers were right. 3D printing has, no doubt, seen some success, however, it hasn't overtaken construction. It hasn't overtaken the home goods market. Or perhaps another one would be VR. VR still hasn't dominated the video games industry. People still buy consoles and PC games. None of us are living in the metaverse either.
So, in that way, I think AI will have a big impact on 2D / concept roles, but not as much for the 3D artist. I gave an example in another comment that I'll paste here to give a specific scenario of what I'm talking about:
"Let me provide a practical example: Company XYZ wants to showcase their new product at a trade show. Everything has already been designed. Prototypes have been made, and the concept portion of the pipeline has been long established. Now, it's the job of the 3D artist to make Company XYZ's dream come true by making a cool reveal animation as the CEO is talking on stage. We have a product that nobody has ever seen before. It needs to look EXACTLY as it is in real life. And, Company XYZ wants to get realllly picky about everything because the CEO has been dreaming of this moment for the past 5 years.
How is AI going to do that? How is the 3D artist's job similar to the 2D artists job? How are you going to train it to create something that's never existed before? What happens when AI gets the dimensions of the product off jusssst a little bit? What tools does the 3D artist have to fix these issues and made sure that each shot has perfect consistency? Again, customization and fine-tuned control is the name of the game for 3D artists, and their work is incomparable to 2D artists.
Even if AI continues to make incredible leaps in progress, it's not going to impact 3D artists in the same way as 2D artists because it's a whole different world of considerations."
Level headed analysis? Yes. Correct assessment of the future? No.
Well, but... why? ¯\_(ツ)_/¯
@@cgforge computers are increasing in computations per second linearly. Computers are projected to overtake the capability of human kind by 2030.
After watching this video I pray you have a relationship with God and Jesus Christ as your lord and savior. If you dont well then I hope this comment at least opens your heart and mind to God one day❤🙏🏽