AI

Really ? If I were you i would quit creating 3d models forever.

very weak Ai
I used a 3D Ai website to generate sniper but it seems something like tree stems XD

For those who think 3D models are still needed in near future, consider the following.
Recently OpenAI showcased a new AI model (Sora) that can produce photo-real video’s.
For now the frames of the video’s take around a second to render, but soon AI models will probably render at 60 frames a second. This cannot jet be done on typical computers, but behind the scenes multiple companies are working on neuromorphic chips that wilt be very capable of doing this locally on a regular computer (rumored even on next gen standalone VR headsets).

Note: 3D polygonal models exist to get the correct pixel colors we need in the frames we render for video or gameplay (for now).

The next types of AI models like Sora will basically just do that (give correct pixels), but way better. They will understand perfectly how physics works, the understanding of light distribution, reflections, shadows, water, etc.

In short, we will no longer need 3D polygon models to get/render the correct pixels.

We are heading (fast) towards artificial brains that will render worlds on the fly (photo-real).
Game engines like Unreal engine will probably going to function like some sort of place holders for AI meta data (Prompt containers). AI models in combination with supper fast neuromorphic chips (GPU analog/replacements), will then render everything on the fly.

Imagine in a game you would slice some cheese, how would you do that with a 3D polygon model?
Or how do you render a realistic waterfall with 3D polygon models?

AI will all do this because it just knows how all those pixels need to be rendered correctly to get the perfect effect.

It will be the end of 3D as we know it.

We will probably still create 3D models, but in totally different way as we do it right now.
My best guess is, we will be showing images, sketches and talk to the AI, then things will be generated on the fly and be stored in memory (no geometric 3D model), it will be some token or seed that then represents the object (can then be stored in token placeholder in a game engine or some sorts).

Take everything I say here with a grain of salt, but this is what I believe we are heading towards very fast.

In addition NVIDIA has released a new Blackwell GPU architecture that should significantly improve AI computing performance and announced Project GR00T (Robotic Foundation Model). But despite all this AI is distinguished from the results of the work of a living designer controlled by an ordinary boring brain without advanced chips by too much variability and inconstancy so it can only be used at the initial development stage, as generator ideas and sketches. AI is not able to offer a 100% accurate result, no matter how accurate a prompt some kind of "creative" is always added. If AI learns to do everything with any set accuracy and acquires long-term mobile memory, one question will arise: why does AI need human population. Question about designers will be ranked 100th in importance.

I think AI generating anything complex is many years away. Output from AI is skewed when only its successes are shown. I’ve seen screwups and embarrassingly odd solutions to problem AI has been given to solve. Which is little surprise to me, since imperfect beings created the programs, apps and software that AI uses. If I would be given a thousand chances making something useful I would eventually do it as well. But AI can, of course, make the mistakes faster than I could. Eventually it will learn to do things properly but under our direction….An imperfect being will have to write the code to fix that. So inherently it will always be flawed.

I've been in this industry for coming up to 20 years now, and throughout that entire time this doom and gloom talk was always present, if it's not this thing, then it's something else. But here we are 20 years later, still going. My point is, none of us know just how this is going to turn out. Admittedly, the AI threat is gloomier than anything last 20 years, but my point stands. If it's stressing you so much, then get out of the industry and do something else; I'm just going to sit here and keep at it and just see how it all turn out. CG is not my only source of income, as I imagine is the same for pretty much everyone else here ... and if it is your only source of income, then I guess you have bigger things to worry about. Worst comes to worst, CG will make me a little beer money each week.

One of my associates is a developer in the AI industry, and according to him, it's not as nice and clean as they want you to believe. AI has a lot of misses, something like a 1,000 failures for every nice image/video/whatever you see in the media. Of course it will get better over time, but just don't believe the media hype, it's propaganda.

Also, let's not forget that companies producing AI aren't do-gooders doing it for the betterment of mankind. They're profit driven, and greedy. So when AI matures enough, you bet they'll charge an arm and a leg to use it.

So I say again, none of us know how this is going to turn out. Or maybe I'm burying my head in the sand ... dunno ... I guess we'll find out in time.