The artistic community, do we have to be afraid of artificial intelligence?
First of all, I don't know if I'm posting this topic in the right place, if necessary, a moderator can move this where it's convenient.
Hello to the whole community.
We are living a moment in the history of technological evolution where artificial intelligence is taking the lead in almost all fields of work. And how not, it is also sticking its head out in human activity by excellence: ART and CREATIVITY.
From text generated after a handful of inputs, to images that mimic any existing style generated from words; infinite music generated from the selection of a few options, programming from a vague instruction, 3D modeling from one or several photographies, auto video generation, realistic animation without the need for actors, and realistic voice, to mention some examples.
To tell the truth, I'm really afraid of what will happen to most of us in the next few years.
Look at this new tool, just with a few photos as input, it generates 3D models with a very good mesh. For now, access to this AI is expensive, but I'm sure that competition and open source software will not take long to appear (as it has already happened with image generation systems like Stable Diffusion).
It is good that an artificial intelligence provides guidelines for modeling/drawing, or do boring tasks like secondary background models. Up to this point, I would not see much of a problem. This way the artificial intelligence would give us more time to dedicate ourselves to the important parts of projects and achieve a higher level of quality.
I would like to know what is your opinion on this subject. What position are we all going to take when artificial intelligence in 3D modeling is standardized and working in the majority of videogames, animation and VFX development companies being as talented as a junior artist, or even superior?
In my humble opinion, artificial intelligence could exist to solve real problems for people's lives, free us from hard, boring and repetitive work. Or helping in research against diseases, data analysis, exploration of the universe,... I think the mentioned sectors are where artificial intelligences can provide a noble and beneficial task for humanity.
The fact that an artificial intelligence eliminates the opportunity for humans to enter the world of art, whatever type it is, erases the basic human pillar. Art is what differentiates us from any other thinking being in the world, art is an expression of human feelings and emotions, embodied on any creative medium.
What feelings can a handful of algorithms capture? It can't capture anything, it can only copy feelings in the form of art and create variations of it.
And I wonder, if in 5 years this technology is going to be as talented as any junior artist, is it worth investing thousands of dollars to learn 3D modeling? Are companies going to have room for junior artists?
I have talked about 3D mainly because it is what I do, but everything said here, can be transferred to 2D concept art, junior programmers, musicians, voice actors, animators, animation actors, screenwriters, level designers .... And I'm sure I'm leaving out many others.
Replies
Kaedim is believed to be an outsourcing scam. They might use some tricks to help generate a model from an image, but there's a lot of messy modeling done by hand, likely by cheap out sourced labor very quickly.
Same discussions cropped up when photogrammetry became a thing.
My personal take is that there will always be a drive to automate but there will never be a magic button for (good) art.
Also, some form of procedural generation (PDG) was possible before "AI" but the results are mixed and from what ive seen the best results are when you mix it with custom (hand) made content. I think that this will be the case with AI generated content as well.
In conclusion, there will always be a need for real artists, at least to create stuff that an AI can copy.
I still miss a source for this claim. Do you have any valid link to information?
Mostly people looking at the results, trying to find any similarities or patterns, noticing they look hand modeled by novices more than anything else. The owner made a response to these claims saying they do have people manually insuring quality and cleaning up models, but denies allegations that it isn't mostly AI generated results. There's no consistent behavior how it handles edge loops, if geometry is closed or open, etc.
Also: there's been job posting from the company looking for people to make models in 15 minutes.
I see. So it's speculation. Well, geometry has to be in the one or another way. And consistency is where AI has its biggest problems. I see this at Stable Diffusion. So i remain sceptical that it is all hand made. Hand made would be consistent then. That the models still needs cleaning up afterwards also fits to what the image generating AI like Midjourney or SD does. And for that you need of course people.