you write prompt and the program creates a 3d model merged with other 3d models retrieved from a database. hair particles, sfx, environments, anatomy styles with big muscles or no muscles. Taking images and making the 3d model match all the proportions. writing down a texture map and then it finds it for you and creates an ai generated grunge wall totally original for you to use.
Wishful thinking, controversial issues aside with copyright, how close is this program to being a reality?
Nvidia showed something like this in their Siggraph 2022 video. You spoke to a Virtual Avatar who was capable of generating a 3D environment in front of you. And you could make special requests like adding or removing trees.
I have no idea how this will translate to characters, but it’s very similar.
You could also look into research Epic Games is doing with Unreal Engine. Especially with Metahumans, although it’s only focused on realism for now.
It is already there. A bit pricey though …
I have seen people using Midjourney or DallE to create the concepts for it already. And Stable Diffusion is open source. So here you have your text to 3d model pipeline 
It’s moving there, but not as a single huge leap.
For example, materials are generated with substance, metahuman is already used in realistic games, making anatomy redundant for the most part. Some concept programs exist for generating ideas, and concept designers use them.
AI can’t yet make a model from scratch, which could compete with human input, It can’t process the idea from the start to finish (I mean a production-quality textured model), so all those pieces have to be assembled. Also it doesn’t know a mediocre result from an outstanding one.
But almost each link in the chain has it to some extent, compared to 15 years ago, when everything was relied on sole talent.
It only means we will output the results faster with more versatility. Also, bear in mind, models do outdate, so most models, produced 10 years ago, look very basic. So there will be work for artists.
Artists always move into the area, where AI/technology can’t yet compete, and everything which can be generated, is generated to save the cost.
But it’s not only AI which is biting off the artist’s salaries. It’s also the technology, as happened with photography and video shooting. People don’t need landscapes and portraits, they make their own, and enjoy the creative involvement.
And I even can’t advise going into manual labor, as it’s being replaced as well. We still have humans on those jobs, but I think not for too long. We might find our world look like an abandoned place, quite a lonely picture with long shiny stalls, voice inputs everywhere, everything is interlinked and cozy, but we might feel lonely without human interaction.
So the vision of creating your own worlds with just voice input is not that far-fetched in the long run. It might happen, and I find it captivating.
In the end, we all wanted to be creative, when entered this field, not crunching polygons or dealing with uv;s. Or struggling with “realism”, which is present in reality for granted.
AI itself is much bigger than that. It will never stop learning and improving, so once it reaches 160 IQ, it will go much higher and perhaps at exponential rate, which not a single human can.
I have this bizarre idea that we might creating something bigger than us, and perhaps it’s our goal, to breed a new kind of existence. It actually might be eternal compared to us, as such a powerful mind woudn’t have a problem replicating itself, being armed with technologies humans don’t have at disposal.
All true and all interesting, but in the end, it is always humans that have to input the data, evaluate the results, etc. What “looks good” is always subjective, from the viewpoint of the living entity looking at it. A machine cannot, and never will, be the creative spark. That is the purview of the living being. So, even if AI gets to where it can create amazing 3D scenes, it will require someone to tell it what to create and decide when its results are “good.” That is where the artist will always be needed. Even if you are just telling the machine what you want, it is your vision and creation. The machine is the tool, the artist is the spark.
It’s not far fetch to think this. It’s in fact, an impending doom or reality we face as a species.
Movies like The Matrix or I,Robot demonstrate that once you can mass produce machines with super human qualities, then they’ll seize the opportunity to either overthrow humanity or simply exterminate them.
I’ve always been skeptical of technology. For me, the perfect balance already existed between 1990 ~ 2005. Everything after that has been more detrimental to our existence. Such as smartphones and social media.
If we feed AI with all color combinations, tonal ratio, main colors and accents of top artists from Artstation, can it output predictably good color schemes? I bet it can.
A lot of stuff can be generated. Pixar/Disney-style characters are cute to the majority, and can definitely be generated, based on previous thousands of samples.
The problem is there are not too many artists with the creative spark, the majority are just doers, making an average work, but sufficient to the client. They outperform AI by cumulative combination and flexibility of human input. But you can see, that it can generate models based on text pretty efficiently. So why can’t it actually comprehend such basic commands, as “take all border vertices of the uvw, and align to the horizontal line, where they meet on one loop”? Seems pretty realistic advance of scripting.
I think AI can learn some artistic principles, but the human will evaluate and pick the best results. So the art might become more like photography, where one picks the best, and here lies the artistic merit.
It’s not yet capable of doing that, but it doesn’t mean it won’t be in the future.
The problem with progress it doesn’t stop, and no-one can predict where it leads. So we might open pandora’s box any moment.
I think your assessment of 1990-2006 timespan is a bit biased, as nowadays there are many convenient services, and computers are so much better. Or maybe I didn’t quite get your point. Could you elaborate, please?
I disagree that progress can’t stop. It’s in fact possible, to limit or control our own consumption. An extreme example might be to look at the Amish/Mennonite communities, who still use horses over cars. But to them, life is normal.
As for my assessment for why 1990 ~ 2006 was best? Well, truth be told, a lot of things manage to peak in that era, that we had to go “backwards” to compete.
20 years ago, I could buy a physical copy of Photoshop or 3DS Max, and NEVER worry about being connected to the internet to use them. Today? Both products now force an online subscription, making it impossible to own [legally].
The same is true with video games and even movies. You could buy hard copies of VHS tapes or DVDs, and enjoy your products however you wanted. But now there are games or movies that can only be played by streaming them. If the Online Servers ever go down, so does the product with it forever.
This is when convenience becomes inconvenience. But not only that, it’s absolutely creating a system where only the most intelligent or wealthy people get to call the shots. If in the future, everything we ever purchase requires an online subscription, guess who has all the power in society? The guys running Server farms with all your information on it.
Maybe it’s even possible this type of technology could turn into digital blackmail, since after all, we keep feeding these computers all our life history, that someone can just quickly look it up and begin to make threats knowing exactly where you live.
Again, I’m not saying we should completely cut off technology like the Amish, but we also shouldn’t be rushing to live in a Terminator/Matrix future, where your right to eat food or breathe oxygen are all being monitored by a robot.
Fair points, JordanVFX.
Indeed, convenience interconnection gives has another side, which is a complete openness, and it’s all good and well until everything is good. But once you want to go in the shade, you’re screwed. I mean, try to get rid of your info if you really want from the net quckly, and you can’t. It’s everywhere. People underestimate back sides of transparency, because they hope everything will be fine, and bad people won’t take the advantage of it.
Those subscription models really irritate me, as I turn down purchasing many programs due to this notion of constant financial drain. I want to pay one time a fair price, not indefinitely, because it’s always more.
Incidentally, I think A.I could also be used for a force for good, but ONLY if we neuter it in places so that it wont cause havoc on society.
For example, it has always been a passion of mine to draw cartoon characters, but I’m extremely paranoid or sensitive when it comes to anatomy. Even after studying and buying art books and practicing, I still have fears of making any anatomy mistake.
However, if a robot could look at artwork, and offer pixel perfect critique that puts you on the right path or feel more confident in your skills, that would benefit the human and not tank our industry.
But if we just let robots do all the art for us, because it reaches a point where it can copy anything on the internet and spit back an exact replica, then it’s going to become chaos when millions of people who were already struggling to eat or pay bills are now left with no income at all.
Some say the only real solution to this is to offer Universal Basic Income., and while that sounds like an ok solution in the short term, in the long term, it also means there’s no more competition. Only the government would be allowed to decide who can own more wealth, and that’s where corruption comes into place.
I think such “advisors” already exist, but they will become more sophisticated. Similar to the Upgrade chip AI movie, where it communicates with you, has quite meaningful ramifications of options, and is very long-sighted.
I really need such a thing. It would help with my planning a lot, would be both a psychologist, a friend, a doctor, a teacher, all in one. You could give it appearance you like, akin to Blade Runner Joi. We are already accustomed to speaking with virtual people we’ve never seen, which is by itself ibizarre.
I think Google and others already have fascinating technologies, which are beyond of what we can imagine.
Honestly, no-one cares about weak ones, and human problems won’t be tolerated at all. If an artist become redundant, it’s his/her problems. But I’m sure other kinds of jobs will emerge we yet can’t render. Orchestrating art, for example, you could sing a tune, and it instantly becomes arranged. Or you just say something, and it creates stories. Art by itself will go a long way with such advances, I think virtual worlds will become very charming and captivating. People by nature want to escape reality, otherwise they wouldn’t get into stories, movies, and videogames. It’s an interesting phenomenon to scrutinize.
I don’t see any possibility to limit such technologies. But, from another point, facial recognition is not yet mass-available, or using deepfakes didn’t become a thing anyone can easily use.
I’ve thought about the idea “what if robots creates more jobs?” however, if the technology already exists to replace Artists, why wouldn’t it also replace every other job after that?
There’s already A.I being made that is meant to solve programming. Or even robots that could publish their own science paper.
At some point, it’s just going to devolve into a race to the bottom. If every Artist got fired tomorrow, they’re going to quickly flood the next industry to take them in. But doing so means wages go down, job security is non-existent. And when that industry gets automated, mass layoffs happen again, and the cycle repeats until the last job is gone or secured by robots.
I see the phenomenon being very similar to Global Warming. Burning Greenhouse gases is necessary for civilization to function, but it also means we’re slowly killing ourselves due to pollution. It’s radical to try and ban all cars today, but it’s also radical to create more of them. There has to be a balance so we don’t go extinct from our own technology.
Humans seem to be replaced where it’s possible, it’s just a matter of cost. For example, baristas (people making coffee) can easily be replaced with machines, because you’d just have to create an algorithm which would recognize simple functions like preferred dilution, temperature, proportions, so it could make it the way you want. There are lots of people in Ukraine on those jobs. It’s nice, as you communicate with them, they wish you a “good day”, and other psychological tricks to keep you hooked. But they are definitely replaceable, as well as cashiers on stalls. We have about half of autostalls, and half human-based. But I mostly use automatic, and it’s much faster and convenient. The interaction is not always pleasant. So that’s the question, what people will prefer - dealing with humans or an algorithm. As people are not always nice, but algorithm is always predictable. It always speaks with you by a young “attractive” female, which is… eer, ok I won’t debate on this. It just means people prefer it.
You think the choice will be made on the upper level, but I think it will be made by market itself. People will have a choice, if they want something AI-based, or human-based. It can be a new kind of production, labeled as such. Mass-produced, craft-made, AI-made. AI development can easily find ways to snooze social upheavals by generous grants to medicine, science or charity.
What you’ve mentioned are just small ridges on the horizon of events, which some people find funny, but it will sweep the society as we know it today. Boston Dynamics might seem awkward now, but with time it will mature, and once it surpasses humans in its collateral mechanics, oh god, it will be a horrendous weapon.
Yes, but it will not know they are good. That will take human evaluation. Also, they will not be able to innovate or come up with something new. All they will be able to do is put out some combination or “change up” of something already done, as all they will have to go on is what has already been done. That’s what I mean by the creative spark. The leaps of logic, intuition, and straight up creativity are only made by living organisms.
Boys and girls, the technology is here. And its not even as complex as i thought. Ai compositor for blender addon turns 3d images into any style you want and added prompts, it wont be long before an animation feature will arrive, and we can all make high resolution movies in are laptops .
Wherever it is practicable, replacing humans with other species seems to be the cheaper option. For instance, baristas (people who prepare coffee) may easily be replaced by machines , if an algorithm could be developed that could detect basic functions like preferred dilution, temperature, and quantities, allowing the coffee to be made as desired.
As an industry dude (Ha, I can finally use that term) I’m still a bit skeptical of the tech.
Not because I doubt its potential or usefulness, but because there exists a certain controversy around how these AI models were trained. In particular, the issue of copyright. Music generators were specifically told to not used licensed music, whereas Art A.I didn’t seem to care where the input source was coming from (i.e scraping the web of any artist portfolio).
No other art tool has that stigma behind it, and in a production environment that could lead to legal trouble.
That said, I won’t live in denial if it can pass the ethics test. Just hope that all this dependence on better tech doesn’t lead us into a Matrix/Terminator future, where it can enslave us…
I think the artist should be compensated. for every time a person generates an image using an artist name. A royalty ( even a cent can add up to a lot) should be given to the artist. think about someone rendering a movie the style of name your artist ____ and then that artist paid for every frame. I’m all for artist rights to defend their hard work. i think just how youtube came around to paying artist their ad revenue for each of their songs, we need something like that for visual artist and also give them the option to opt out and banned their name from the prompt library. They already ban the word sex and nudity. names should also follow.
I don’t use names of artist, my rule is only classical artist who are long dead.
The thing that blew my mind about the prompt, was that it was able to decern the sheep vest from the hair. I specifically wrote down to make the vest out of sheep hair and it knew exactly to locate the vest and apply a better-looking hair particle than the one I have, that really changes the whole game and how we textures things. even turning low poly models into highly resolution models with a compositor.
