I wasted lots and lots of time this year trying to find anything useful for content creation. Models/textures/ decals/ just nice alpha for brushes .
Contrary to all that internet buzz and doom scenarios I found really nothing . All AI can generate is maybe some phone games fancy art : 2d pictures for a fantasy style "find a thing " game or maybe super cheap looking 3d somehting. A thing you would do with a pistol to you head and tomorrow deadline. Nothing that would realty help you to be more competitive on a job market. Have an impression a generative AI is only good for monetizing youtube shorts.
Totally other story with scripting AI . Especially for Photoshop and Blender. I use mostly chat GPT . It really elevates you interaction with Blender and now Photoshop up to totally new level. Still I can't get anything useful from GPT for 3d max for example, or Designer /Painter.
What about other AIs ? Does anyone have an experience to share ?
Replies
https://slashdot.org/story/25/10/02/191224/ai-has-already-run-out-of-training-data-goldmans-data-chief-says
..the next thing is to "imagine" something.. this remind me of people on LSD to make.. art.. ??
So better check when "generating" for ecample a mouse you do not get some letter from the lawyers of Walt D. .
Or a duck, a hound, bear or whatever..
if you're genuinely interested in getting useful ML based tools to help with things you need to do - you have to train them yourself.
Nobody is investing in making anything useful because there's so much money to be made scamming investors that it makes no sense to do anything other than claim your gpt wrapper web-app will change the path of humanity and use the bags of money people hand you to buy a gold yacht full of cocaine and dick-pills.
when the arse falls out of it all we'll start to see something useful that we can buy - it'll be a couple more years at least though
That's bad. A generation of people who can't think for themselves. But that's what they want and here we are.
There was this anecdote (true or not.. does not really matter) when a woman ask Picasso for a small drawing and he did a quick on one on a napkin and presented this with the (somehtink like this) words : "Here we are and you can have this for 100000 franc." The woman said: "What !! But you needed for this just two minutes." And he answered: "But you pay me not for this two minutes but for the yearss i needed to be able to do so."
I'm not sure if he (or should have) added: "But you can try for yourself."
So back to the "reasoning". AI may be helpful if trained on one own data.. but then: if one can not recreate something similar as one already did then one might be a long step behind calling oneself artisan nor artist. But one properly can not argue with those users who simple collage something they got somewhere.
I also to not agree that "we" need this.. because humankind did alway made amazing things looonnngg before the first computer was build and there are people who do make this even today without using a computer even if the may make their work puplic via a computer. So i think "those" AI is mostly needed by uncreative people and i include developers because i know something about this area.. programming is also some kind of art.. and also do not forget writers...
(typed almost in a rage and not really spellchecked.. also only using MI.. my ( little bit of ) intelligence
Also, the AI added to UE5 is a dumb chatbot, you're better off using Google Gemini or ChatGPT.
It's really nice for formulas, general algorithms in natural language or small isolated chunks of code, but trying to use it for complete script (especially something complex) is unreliable. Sometimes it will work out of the box, other times you would need to fix some/lots of it, mostly it will hallucinate and call nonExistentStuffThatDoesExactlyWhatYouWant() functions.
Modeling ones become better. Still not ready to be used for production and there are legal issues with that anyways, but you can use it as a base for sculpting or a guide - it's much easier and faster to model "over" something when you already have proper dimension, proportions, important details.
Movie Studio Lionsgate is Struggling to Make AI-Generated Films With Runway - https://petapixel.com/2025/09/23/movie-studio-lionsgate-is-struggling-to-make-ai-generated-films-with-runway/
Elon Musk’s xAI hiring video game developers - https://www.foxbusiness.com/fox-news-tech/elon-musks-xai-hiring-video-game-developers
I always wondered why people does have more faith in "a big" company or person "saying" that they will do "the next big thing" when they already often have proved that they failed in serveral things, just because they made "one big hit" (or two), and so made those company or person bigger than before so that they can "turn down" those who already does a lot of things in those areas/domain they claim to be superior (with no real evidence).. somewhere in the future.. with some delay.. maybe later.. just wait for it.. we recently bought this company ( because their value was falling because of our promises we could not satisfy.. ever..) and we are thrilled to share this news which elevate the user experience... and by the way: it's now subscribtion based.. and because of the soooo much higher value and include templates just slighlty more expensive than the other things.. but you can save a big deal when subscribing to the XL bundle to replace all the other applications you are already using and used to.
</sarcasm>
So if you want a tool that isnt meant to cut out the entire journey and jump straight to the "end result" (very much debatable) you will have to train your own.
Like, make me good UVs, with seams in the right places, rectilinear layout, minimal stretch, overlaps for reused bits, etc.
But for it to solve that, it would need a pile of good source material to train on. And that's the problem.
Also if any future artist never learn anything about art but just tell some AI what to do.. the next thing will be that the producer directly talk to any AI..
..so: Where is the benefit for any artist in the long term ?