I Lost My 3D Artist Job Through Midjourney 

handcraft is now one of the best payed jobs, cuz nobody want to do it and everybody needs one

Eventually they will understand!

@@Prajwal____ One hopes so, but I know how greed tends to take over!

AI does not regurgitate the training data as is but mixes it up according to the prompt to create the request image. That means it can technically produce infinite combinations of the billions of images it has seen. A human lifetime will not be enough for someone to sit in front of MidJourney and make the valid claim that they have seen it all. Also, These technologies are at their beginning and are improving very fast so I wouldn't bet on them reaching their limits.

@@Kevin-jc1fx True, but it creates nothing truly original. It is based on things others have done and it is going to be used against the average man sooner rather than later! Hell, Levi's has already switched to computer-generated models for their clothing and has gotten rid of their real models! This will be the final dissolution of our society if it continues at this pace!

It's mean they wil just repeat making the same product. Yeah just "same mass product" not new one.

White collar jobs too, doctors and lawyers are getting replaced by AI in some areas. CEOs will be replaced too once a (broad)AI is out that can employ other (narrow)AI dynamically as needed to create and run a business idea. What one think won't be replaced by AI will be the very thing that AI scientist are working to replace because that is where the challenge lies not in packing boxes or stitching clothes. The best way to speed up AI development is to raise minimum wages so companies will be forced to use AI/robots.

Tools like ControlNet have made diffusion models a lot more easy to steer but all those are still at their beginning. There are already stunning results from text-to-video models too. A few years from now, the progress will be mind blowing. I don't see a way for this to be stopped. Everyone should try and figure out where they can fit in that context because proper regulation will not be guaranteed.

That is good analogy. Automated process is more efficient(in most cases) but handmade product is much better quality. What I believe we will see is AI being used as mass production tool and artists with passion will switch more into one time/comission type of making products.

@@dexterdexter7923 I disagree with that deduction, because the automation in previous century WAS the mass production tool that you mentioned. So lot of white collar workers can get away with selling their handmade high quality "One time/commission type" product (i.e. web developer, artist, 3D modeler, musician, film director, etc.) The AI we have now is not only mass production tool but also producer of "one time/commission type" product. This was why ChatGPT, Latent diffusion, AlphaGO, etc. are so revolutionary. They not only mass produce but also create unique product for each result they spit out. Whether it be a new song, writings, art, website, or even new app. All are tailored made according to user's input or prompt. PS: This was why a lot of white collar workers are in panic right now. Especially the newly graduated students who spent a quarter of their life to pursue career which are being eclipsed by AI.

That is so very true. I, as an ex 3D artist, got bored of the repeating part of 3D creation (UVs and retop mainly). AI could have been sooooo beneficial. It's crazy to see that all the efforts are going to the prompt market which is not even remotly fun for artist who can never end up exactly with what they had in their mind...

All artists study photographs and other artist's works, whether it be old masters or browsing through Artstation and try to find ideas or even trace/replicate the artwork without artist's consent. AI also learn art that way too. AI pick out the features from the piece, not copying the whole file, then process it and store into database aka its brain. Similar to how artist took another artwork and try to outline all the shapes, measure the proportion, and model the value rather than copy down inch by inch. These AI are mimicking human's learning habit basically. It's clear what AI is trying to replace the artist not being artist's assistant, so you have to brace for it whether you like it or not. IMO, I find most complains around this subject very ironic, since all the automations that we have in our daily lives were the one that caused extinction of many labor jobs. For example, you would prefer to buy one-time retopology software over constantly hiring another low-income guy to do the repetitive retopology task, since it's whole lot cheaper. That low-income guy was one of those labor workers who lost his job to the machine. His net income is now zero, and he is forced to adapt or die. In the same manner, many clients would prefer to buy one-time creative AI artist over constantly hiring artist to the creative job. So, artists are now forced to adapt to this change. I mean, if it's unethical for AI to invade the creative jobs then it should've also been unethical for machines to invade labor jobs. Both parties did their best for perfect products, one physically and another mentality.

@@qull3840 Don't get me wrong I didn't say it was unethical for ai to be part of the creative field, I just expressed how disappointed I was in the way it is going, towards prompting instead of more profound, useful tools that could actually help. Prompting doesn't make you an artist or an "ai artist", it makes you a commissioner and that's it. You just confirmed what I said in the first place. Technology has always been focused on replacing labor tasks and repetitive processes. Creation and art isn't a labor task nor a repetitive process, retopology is, so why create an ai that does everything? Why would you give up on your creativity? Let the ai do the retopology and not the creative part. Retopology is just an example among thousands of repetitive tasks in the art industry that could be replaced by ai. Also what you described in your first sentence is agi, for the moment we're still in the era of ai and it is not working like a human brain at all.

UV and Retopology are rather difficult to replace with AI. There is still automatic retopology, but it is not used much. Background assets are often used. However, in important character assets, the quality and results are not good. No matter how developed it is, in the end, that part has to go through human post-processing. There is no choice. UV and retopology are more creative and purposeful than you think. However, since there are so many similar parts, such as humanoid characters, it is easy to replace modeling or textures with AI. Eventually, at the end, humans will correct the mistakes made by the AI. It will mainly be the retopology and UV parts.

⁠ … see this are the types of comments that get me going… WHY are people sucking coorps so hard? Do you realize that those jobs are usually entry levels in order for you to make contacts, build relationships, and learn? It’s like an internship? What it’s quite ironic for me, is that you’re ok with something that was inherently created to be a form of human expression, being generated in a push of a button. The lack of empathy for those who took the time to learn proportions, shading, lighting, composition to find a voice of their own and being able to achieve something they’re proud on. Like a track and field runner working hard to go to state championship to be competing against the guy with the scooter. It’s still a race. And he is pressing buttons. 😂 But “AI is art” But let’s see what it takes to make great manga panels shall we? “as a Sui Ishida style, black and white Japanese traditional manga comic sketch drawing, A dark short haired femenine woman wearing a plain white uniform as a volleyball player performing a powerful serve on a match , special effects, full length wide body shot, 4k, volleyball field background, seed” yeah… I see what you mean… 😂 why would you need to know anything about special effects, drawing skills when you can literally throw words. I love the part when people blatantly put the artist names. It’s my favorite. But please tell me more about how wonderful ai is.

the government never makes things better and why? ai does not violate copyright unless prompted to but we wouldn't sue photoshop because someone used it for nefarious reasons, ai is a tool. the person in the video is the same as the people who thought drawing tablets would destroy the art economy because people wont buy art supplies and that art will lose its soul because it doesn't take as much technique.

Government never cares about artists, all those tax incentive in VFX industry already turned artists into modern skillful slaves.

@@stannisthemannis8694L take. People did think digital drawing would get rid of regular art in the same way but this is literally just you not making any decisions in the art itself. It’s like going to the super Parker picking an exact carrot you like and calling yourself an artist - it’s becoming that simple. So yes there is little skill involved.

Lol. What's your source? Your dreams? XD

How do you feel about what's happened in china?

I agree (today for myself and maybe I'm wrong for a lot of other people and the future) because at the core. as unfair as it can be, it has always been about adapting and finding a niche or an individual way to survive on this planet. On the surface it's different, from the beginning on I had the feeling that ai generators don't hand over much control to us artists and I think it is already changing. + At the moment we witness a new technology rising without exactly knowing where it's going to lead. We can assume that there will be a whole ai generated entertainment bubble that is created by just pushing merely a button but still humans are the customers and they like humans doing amazing things. We much more need to show this amazing handmade work and our skills and ability of how we use ai to enhance the quality of our work so people can see, appreciate and pay for it.

Ministry of flat is essentially machine learning UV unwrapping and it has removed 90% of the work from making UVs highly reccomend it.

@@qaztim11 thank you, I'll give it a try)

Same here bro what are u doing now

well its not. the title is the reddit post and the video responds to it

@@stannisthemannis8694 They are different, although similar.

actually good comparison. Its time to jump onto the train

True, tried some prompts, all look the same, same style background, layout, it's uninteresting

@@donkeykong6626 Yep and they all have similar flaws too.

Some of the most imaginative car designs I've ever seen have come from Midjourney.

@@seankelly5318 They are and only ever will be recombinations of existing images, so if they did not train the AI on some "stolen" artwork or photo of a given tagged topic then you will never be able to create variations on it. This is literally how these systems work at the coding and data level.

@@DanielSMatthews This is what most designers do for a living, combine what they see into new designs. It's how imaginatively they are combined what is fresh.

Unfortunately, no one is asking our opinion.

Unfortunately, no one is asking our opinion.

@@Kevin-jc1fx I don't have the need for anyone to ask, I still have the right to state my opinion and also to make a fool of myself when necessary.

@@anubismacc8165 That's true, we all have freedom of speech, even if no one listens. 😂 OpenAI Ceo was in congress recently and said that AI research should be urgently regulated. He sounded more like "let's keep others from advancing too fast so that we stay ahead". MidJourney is far ahead for now but Dall-e keeps improving. While talking about regulations, he didn't seem concerned about artists rights unfortunately. We will see how it unfolds

Good luck.

AI will be inspired by your human creation to create highly inspired AI creation.

@@nsevv true some will look at it that way, some will definitely use it that way. Until people start adapting toward the potential of Ai has and understanding generative Ai’s ethical crossroads we are in, people will lose from it.

AI image generator can be several things rolled into 1 package: a search engine (used in conjunction with something like Pinterest), a mass concept art generator (that can generate hundreds of ideas per day), a render engine (no more tedious meticulous shading, through touching up still needed), a converter (it can convert a finished art piece into a different render style, 3d art into 2d or 2d sketch into 3d looking image for example)

If us humans already looked at and consumed information similarly to this technology and used it as inspiration for our own art styles then aren't we also just very inconsistent scrapers as well? Everyone has a style or styles that we tend to gravitate towards at the end of the day. And even with the aid of tools like backwards engineering the perspective to be used in 3D to help with modeling the backgrounds for an illustration we're still subjecting these techniques to the specific art styles we're attempting to draw or paint. The proper way to create an illustration anyways is to do some studies first and then apply the learned knowledge to the new piece you're working on. All this does is skip the studies part and let's you peek into the future to see how it should look. It seems the main problem is frankly the not so moral way these libraries were attained (Even though most art was under public law) as well as us humans losing an ability that for a lot of people was extremely validating. If you're an artist that had a fanbase then even as of 2023 I still see plenty artists earning pretty well from their patreons as well as other commissions. The main hit seems to have come from industry art specifically the illustration side of things. With jobs going down as far as %70. This is due to certain concept art requiring very specific designs that require the human touch ones the "Random" button that is AI seems to not cut quite cut it from multiple angles. There is also a pretty good chance that most concept artists were already 3D modelers anyways to some degree. If you're familiar with the advances in AI for the 3D field you'll know that in general creating a model from text to image is possible but only at organic levels and not complex hard surface due to the sheer amount of subdivision needed in the model to resemble the form. The retopo nearly always ruins the cleanliness of hardsurface if done at an automatic level. meaning manual retopo and uv unwrapping is still very difficult for AI in which case it's best to model still from scratch correctly these hard surfaces instead. On the VFX side most AI advancements have to do with camera tracking , mocap or facial deepfaking as well as compositing. For some reason there is still a really massive hole in the process of creating successful proper working particles systems. Since these workflows vary widely from studio to studio and considerably alter the documented methods. Simulations remain relatively untouched by full AI automation even on the more basic maths and pseudo programming involved with it like VEX. Oddly enough if you continue up the ladder still on to the programming side of things which involve more relatively well documented languages like python or C++ border on full automation. Now anyone with an IDE can just work in tandem with AI to achieve results and learn 10 times faster than ever before , even if you've never programmed something similar to what you're doing. Way at the front end of academia all math degrees and compsci are threatened as well. Seems the only way to stay safe for longer is to know obscure knowledges that AI simply cannot find its way on to. Though it's only a matter of time before it starts leaking from in house studio staff giving away context info for it or the actual studios signing specialized learning monitoring agreements with OpenAi in which every action is studied and documented by the AI. Simply put there is no hiding from this tech.

Please Elaborate :)

I didn't understand that either and would like to know how that is possible.

Hollow Knight

Abruti.

I know most people don't want to hear this, but here goes: That's an idealistic answer. Not everyone can make it in the field, and people need to make hard decisions to survive. A lot of people don't get to choose what they want to do simply because they want to do it, as you grow older you'll understand. I'm sure a lot of people wanted to become something else at one point in their lives, but they are here now. If you're not looking to get paid, then it's never gonna be a problem. In other words, as a hobbyist, it's not gonna be a problem. In fact, these AI tools can only add and not take away from a hobbyist. People who need to put food on the table need to think about it and cannot be idealists. If you're just starting out and you do want this as a career, not as a hobby, then think carefully. Some jobs are just gone for good, and others will follow. Choosing to become a concept artist at such a time is just plain stupid, for example.

This is an overly optimistic view. Many people get into the field because the love the process and push towards mastery. Drawing is much more engaging than using prompts. Prompts turn something artistic and engaging into something that feels like homework. It's like if you like woodworking and carpentry and turn into a guy that just loads up files into a CNC rig and places the wood in the slab...then hit's a button.

This is only answer, my friend

@@sp4cedo9 How is it even you doing 3D though if all you do is use a prompt library? When I order food at the drive through, it doesn't make me the cook.

you’re right, AI isn’t a problem. Ppl like you is.

@@arturasba7387 exactly

@@reeanimationgaming1034 you could create nfts from your pride only xD

@@arturasba7387 What do you mean?

@@reeanimationgaming1034 you value only the end product and ignore the processes that allow it to come into being. If you allowed yourself to doubt for a moment that AI is not "holy progress" (if it is progress at all) justifying the genocide of creative people jobs (concept artists, etc), then you wouldn't sound like a blind man teaching about color theory.

When you say that I lost my job, I did not understand that artificial intelligence creates 2D images, not 3D?? at that time. blender Rhino maya 3dmax ................ in the company and close it??

if hes doom and gloom your a bit too bullish imho. reality is always somewhere in-between.

@@marsmotion what did he say that was bullshit?

@@stannisthemannis8694 yes bullshit : )

@@GameBoyyearsago you're cringe

@@stannisthemannis8694 so as you are cringe too : )

and the pen, phone and laptops.

A ton were created though after. You need tons of people to make and service them. The problem with A.I tools is you need SIGNIFICANTLY less people in that field as the A.I can start to be used in the process of building itself and servicing itself.

There was no property theft behind those inventions. A.I companies are stealing the images of thousands of artists for their own benefit. That´s intellectual property theft

Trees and Mountains are beautiful work of art created by nature.

@@nsevvdepends on who you ask