CG Geek vs. Ai
ฝัง
- เผยแพร่เมื่อ 9 พ.ค. 2024
- CG Geek vs. Ai in a Blender 3D art competition!
Check out the ROG Zephyrus G14 laptop: bit.ly/CGGEEK
The Blender Nature Essentials: blendermarket.com/products/na...
patreon: / cggeek
OpenVDB cloud pack: jangafx.com/software/embergen...
Blender Cloud Material Tutorial: • Most CG Clouds Aren't ...
CG Geeks Discord! - / discord
Music: share.epidemicsound.com/cggeek click this link for a free month!
subscribble for more giggle: th-cam.com/users/Blenderf...
See ya over on da grams: / cg_geeks
or if ur a tweeter: / cggeeks
Thanks to ROG for sponsoring this video
#3danimation #ROGZephyrusG14 #Transcendence #ad
Who do you think won?! Happy to be back, needed a few months to recharge, but ready to go now! ~stay tuned for more content coming soon..
Edit: I say Midjourney in the video -but it appears that I'm using a generic stable diffusion Ai art generation. My bad! You can see some of the Midjourney results on the discord server here: discord.com/invite/awfcPvG
Bring video on steps to learn graphic design for beginners pls
I think your image is better. I also agree with you about the detractor to imagination by using AI. If one relies too much on AI, they are no longer creating art from their own imagination, the AI is creating.
New VFX idea, Make a VFX tutorial snow boarding a mountain when it explodes.
🔥🔥🔥🔥
Your image is DEFINITELY better.
It's the same thing in my fields - programming and music making.
AI is good for fillers, for saving time, for those mundane tasks you always have to go through in every project.
But the technology is just not there at the moment to completely replace programmers or artists.
because you type prompts wrong. You have to explain it like to a 5 year old.
/imagine a floating stone structure floating in the skies. Moss and alien plants growing beside a path in the stone structur. Misty, foggy environment with a deep perspective. A distant futuristic aircraft is passing by. Two young man, full body, standing with the back facing the camera enjoying the perspective view. Mystic scene, floating, ancient, alien, photo realistic, natural colors. (then add some technical stuff).
First try this prompt is already very good.
A floating stone structure floating in the skies. Moss and alien plants growing beside a path in the stone structur. The stone structure comes in the shape of an infinity symbol. Misty, foggy environment with a deep perspective. A distant futuristic aircraft is passing by. Two young man, full body, standing with the back facing the camera enjoying the perspective view. Mystic scene, floating, ancient, alien, photo realistic, natural colors.
This one works even better, it even gets it with the infinity symbol shaped stone structure.
I work in 3D production and AI has basically become standard practice here. It's mostly used for reference material, concept art and resource gathering, but the artists are still very much required and everything in the final products are still hand-made by real people.
AI is basically a tool, not competition
For now
@@user-cs8yj1jw1f and hopefully forever, fuck AI
while it's impressive, and how quickly it evolves, from personal experience in the industry, it's going to take quite a while before it can actually REPLACE people.
Our work may have changed and adapted, but we'll still need people to be there for a while
@@SharpWind It'll never replace *all* the people. But it will replace 99% of them. The company I work for just laid off its entire creative department of copywriters and graphic designers, bar a single person told to use ChatGPT and MidJourney from now on. Team of 5 reduced to 1.
@@nathanrussell-raby5460 thats what im saying. It really wont. None of us have been impacted, and the amount of supervision and manual labor it needs, completely throws away any threat of "AI replacing human artists". Its not nearly on the level, and it wont be for a while
Remember without artist like you ai wouldn’t be able to generate anything
so would you, without your eyes you wont be able to paint anything
@@shujin6600well, that's just silly. For one thing, there are absolutely blind artists out there; Eşref Armağan was literally born without eyes and taught himself to write and paint. For another, obviously human artists don't merely copy the work of other human artists, because there was a first human artist. Human art came into existence spontaneously, and humans create new artistic styles, genres, and techniques on a daily basis. There's absolutely no comparison between the process humans use to create an image and the process an AI uses - you literally just watched a twelve minute video where a human artist discussed the different creative decisions he made in a way that no generative AI model could possibly do, because generative AI models are just using statistical models to find the arrangement of pixels that, when run through an algorithm to embed it as a vector in multi-dimensional space, best matches the embedding of the text prompt.
@@blockalism you're calling me silly while you're saying that there was some who first made art. I know its hard for dumbheads like you to understand but, here listen.
anything which we have even drawn or created is from nature we only mix and match them and NOTHING MORE. I want you to show me an art piece which had nothing from this world. our brain is a storage system not a creation system.
adding more step in creating something doesn't mean being creative. AND there was an experiment done to check creativeness of AI vs human. AI outperformed average humans. (so stop yapping about being creative)
if i say Apple you get picture on an apple inside your head, just like how AI do.
even the first found drawing was of something they saw in their life and at most transformed a lil bit (just like how fucking AI does it)
but if I tell you "orangendamus" you would get orange in your mind which is not called being creative or creating something new you got that image from an already existing thing (orange).
And the bling artist had to first learn the stuff from touching and feeling (which was then stored in his brain and then mapped out) and then he was able to draw. and if you give AI the ability to understand dimension of something from touching he would be able to do stuff as well.
@@blockalismwell said.
@@shujin6600 blind painters exist, try again bud
It pleases me greatly how high performance laptops have been dropping the super gamer cringe look and are looking more and more like normal laptops.
Agreed, love the aesthetic of the Zephyrus G14. Looks professional and sleek at the same time.
Now hopefully their prices come down to normal laptop prices 😂😭
Hey! i remember seeing your blender tutorials! so nice to see you back!❤
Glad to know AI didn't steal your memories!
Great video! As Omi is a 3D solution - for people who doesn't use 3D software - we are also assessing the impact of AI on our business and including it in our approach. That said, we don't believe it can replace artists.
The way I see A.I being used is that it's going to help with initial ideas, like you said, and to help us with the boring/tedious tasks like re-topology, UV unwrapping etc. So we can focus on creating and painting.
It'll be another tool to use to help the workflow become much smoother, quicker and less tedious.
Happy to see you back!
Nice number 1.11 million subscribers!
finally a video from you its been a while
good to see you again
Ever since these AI bots have flourished the internet, I often get comments like "which AI software do you use?"
How easily people label 50+ hours effort as mere typing a prompt
No matter what people say, we CG Geeks know the freedom and precision that AI can't offer (yet).
I feel like every time I watch your videos you keep upping the quality 🙂
Glad you're back! Your videos always motivate me to create and keep me inspired. Keep up the awesome work!
He's back!
*"Draw hand with 5 fingers"*
Ai - Cannot compute
kkkkkkkkkkkkk
i love your stuff dude and its awesome how you're still positive through all this.
Good to see you're still uploading!
This guy was (and still is) my childhood
Your render is amazing and I will always stand behind human-made art, but that was not MidJourney and you can tell, it looks like what Dall-E was generating a year ago
amazing once again :)
by the way how did you uv unwrap the geometry nodes set up? i have done something similar but when i keep drawing meshes on top the material is not always looking right.
I tried to ask ChatGPT the best way to animate one vert of a Bézier curve. It gave me a 14-step instruction list that completely failed at step 11, telling me to add a keyframe to the vertex while in edit mode.
I explain to it that this doesn't work, and it proceeded to give me four other sets of instructions, each of which failed catastrophically worse than the last.
AI sucks at Blender.
I'm also a 3D Artist for many Years and using Blender, ^^ and in my personal Opinion, AI will never fully replace real Artist, on the Surface it may look like it would come close to it, but in reality it is far far of from it, two major points are,
that for one AI is limited in its capabilities through the given Prompt, because it will always try to match it while a real Artist is working with it and things more into it, and the reference it has, where the Problem for the AI is that thousands of people daily use AI to generate stuff that looks relatively similar and the same as also other AI will use these also as reference what will then also influence and mainstream the style, while a Artist can free itsel from a reference and just imagine things,
the second reason is that AI alsway tries to make everything perfect (as far as it can make stuff perfect *cough "hand" cough*), while Humans make Mistakes and even Pro Artist with Years of Experience make little Mistakes while they know how to cover or fix them and that then also gives a piece more character that AI can not and, in my opinion, will never be able to copy, ^^
and while I'm strongly against the use of AI for endproducts, I also go along with the endstatement that AI can be a usefull tool to incoorperate in a workflow for concepting and sketching, without having concept Artist worry for their jobs because even they make concepts for their final concepts ^^
but yeah I find your render also better, even ignoring the fact that the AI has not really got the realistic style, yours has just look and felt more complete and immersive ^^
I liked your version much better. It was just a lot more fun.
The tree island made by the AI did look creative and pretty, but something seems missing. I think if a human artist took that image as a concept, and made some changes with a goal in mind, it would look a lot better.
exactly. most good artists try to add a bit of story telling into their animations/renders to give them some more depth. even though AI tries to replicate it, it comes pretty obvious that it is just spitting stuff out based on the prompts
Its nice to see you back
Less Go my dude Cg geek is backk !! Im so happy to see you back,
Thanks for another interesting and fun video CGGeek!
When the world needed him the most the appeared
Good to see you back!
Yours is 1000x better.
But AI spent 1000x less time on it haha
You could use AI for the concept and then make it better. But I agree, takes kind of the imagination aspect out of it and makes your brain less able to come up with stuff.
Great vid
The fact that a single artist will have to compete with remnant aping millions of Artists combined... is terribly astonishing!
Issue with the prompt might have been “floating island” an island is meant to be a land in the middle of water. By saying “floating” it gets quite confusing lol
Great format 💚
Love you missed u. The laptop i have been managing got destroyed and i haven't watched your content since. Seeing your video and your render gave me goosebumps as usual ❤❤❤
finally back you should post more often
Nice video! Of course yours is better XD.
You pretty much said it yourself: you have way more control over your scene. You were also grounding this scene and giving the elements in it an explanation. Especially when it comes to storytelling and things your characters have to interact with it is getting harder and harder to use AI as a timesaver.
You should watch Feng Zhu´s video about AI where he explains that combining concepts and giving them purpose and function by only using AI is not possible because AI doesn´t THINK about what it´s doing. A fantasy world is fine. But try using only AI to create a steampunk photo camera with believable functionalities, different mounts with a fitting suitcase for a videogame, prop artist or manufacturer. This stuff will take you forever. AI is using its database to "even out" the best results. That´s why the results can have interesting ideas as shown but generally look the same in style and composition and won´t have a logic function to them. And that´s why it epically fails when it comes to create different views of a historically accurate scene like 30s New York city with the accurate layout or clothing and enviroments of aristocrates through the centuries.
However: going the opposite route by moodboarding ideas with AI and then perfection it with your art tool of choice like 3D I think is a really productive way to go. Maybe next video could be you starting with AI and then trying to create a better result from that.
Ngl, I think you won this one. Wonderful video!
oh and BTW I screenshot your final image and past it into Krea looks amazing, especially the vegetation, you should try
they don't want us to think. please everybody say no to ai art!!!! Nothing feels better then completing a vision you imagined by yourself
With my latest project I gave Stable Diffusion XL a shot for conceptual ideas for my scene, it's a Tron: Legacy type thing, and like you mentioned about how blowing it up makes it fall apart, you are totally correct. At first glance, in the 127 images I generated, there's some really sweet stuff, but when I started analyzing it to model it, it's nonsense. It's a bunch of lines dissolving into each other and stuff disappearing and it just isn't useful yet for what I'm asking of it, but I do believe in the possibilities. If it could get to the point that it can create blueprint sheets that are accurate that you have all the views needed to model it that would be just... Incredible.
Great video!
No way, im subbed
Yo I was actually thinking of getting that laptop recently, what a coincidence, is there any discount codes I can get from you for it?
I wish I could do something like this
Glad you're back with the milk! :')
You won of course, no ai can beat human. It will just make human lazy.
I like your image, is a scene si yiu can get many angles and make animations like you did, AI is ok to get an idea, a concept of what you want, but it won't replace artists because some clients are very specific with their requests XD AI can't be that accurate even if you try 1000 prompts (I studied what they call "prompt engeenery") it gives amazing results but they never are exactly as we want it.
It is a great pleasure to see you again after a long period. But you didn't continue your previous video in which you got virus in your PC. (You said CG Geek will come back.)
Incredible work bro! you win this time
Good s*** dude. I learned a lot from you when I first started blender, and I was curious what you thought about AI advancements. I find people just seem to care about the end product, and I imagine soon will be able to put anything in our imaginations, into practice. I fear animators like you and I may be obsolete in the next 10 years, but it's skill, focus, and craft that makes our work worth it. That can't be replaced.
That Rog laptop btw is really cool.❤
Yo when I am early to a video lol 11AM is a good time for me it seems.
Awesome but a bit of a unrealistic comparison, if you know how to work with AI you can get waayyyyyy better result. The only problem indeed is getting full control, having a 3D scene is unbeatable at that
I'd like to see you take this further and show ways that you as an artist can take ai content and use it as a tool to create something original.
A legend has returned
Glad your back :)
The betrayal to that poor cube ;-;
New VFX idea, Make a VFX tutorial snow boarding a mountain when it explodes. 🔥🔥🔥🔥
so cool!
Ai will forever and only be a tool, dont be insecured guys, lets be composed
yea what else would it be? AI wont replace human it will be humans using AI would.
Ain't no way that in this comment section are 3 ytubers that im subbed to
Man, this Stephen Hawkings voice was something else... Also, are you sure that isn't just Stephen Hawkings possessing your computer..
Yours is on its own League, significantly better than any AI can even do. And I kind of thought that as soon as I saw it and was correct
I do like the idea of using it as like a way of getting inspiration.
Love your videos ❤
You should add a disclaimer to that Asus ad that the repairs department is sus!
Colors in AI looks good but the structure is better in blender.
Honestly, the AI image is boring and soulless aka AI "art" pretty neatly summarized.
Like don't get me wrong, as concept art or some quick thing to use however it's fine, but it's absolutely no match to the image you created. That's why AI will never fully replace humans. Because AI, while having technical competency, has absolutely nothing to express (or no "soul", if you wanna be more ephemeral about it).
Awesome job!
AI Wont replace humans, it will be the humans using AI. I dont know how blind, unimaginative and uncreative you have to be to not feel anything from seeing anything. People like would NEVER be able to differentiate between AI and human AI.
the AI art bros can say what they want, but good art done by talented people will always out preform AI "art"
@@Shattered3582 nice joke, you all are already replaced (not being replaced, you already have been )
@@shujin6600 alert ai art bro detected
@@Shattered3582 I have made more 3d art than you have (I'm a 3d freelancer, I'm a game developer, I'm working in an established studio). The only difference is that I'm willing to accept the truth unlike you dumbasses.
Welcome back king
How much time does it take to make a short film in blender
I am new in blender and 3d world. People were discouraging me all the time with the fate of AI. This video inspired me a lot. Thanks. Actually every person is creative and unique from the point of view of thought an imagination process. So Nothing can replace the creativity of a real human in term of Creativity.
This wasn't even a competition 😂, you embarrassed that ai bro, is like a pro UFC fighter against a fetus
One day we need rtx macbooks.
Dont get me wrong, I have the m3 max and its crazy fast with blender, but a high end discrete gpu like an rtx would be something else.
Agreed, nothing beats rtx for realtime rendering
Ai is watching this video 💀
😂😂
its not like the thing he created is TOTALLY new, AI would already have data about this type of thing.
Gscatter isn't working in blender 4.1. any help?
Might need an update, works with Blender 4.0 in my testing
I think yours render adds the dimension feature and AI isnt. I mean if I look at your work I feel the dimension
it lacks the dimension because he was not prompting it.
Ai will only ever be a parlor trick. The problem is ceos fall for it
Ai art will not replace us.the people who use ai will replace us
dude you true we are the makers
You're the best man🙇🏻😺
ai may produce results in seconds but
can it meet needs of clients ? no 🗿
Hmm not sure what's going on with the AI art at the end there. You say it's midjourney, but the results you're getting out of it looking nothing like Midjourney. Or at least it resembles V1 of midjourney from 2021. Kinda odd
Just clarified! Missed that I was using a generic stable diffusion instead of Midjourney.. seems like we might need a round 2
@@CGGeek Ah okay I thought that might have been stable diffusion! Great job on the video anyways!
yeah whatever stable diffusion thing you were using was crap compared to what AI can do nowadays ngl. love your piece tho, just seemed unhonest a bit. i see your pinned comment tho
Back at it Sensei!! GG Geek Wins
the main problem with AI generation is the non reproductivity (because of the randomness) and even with tones of controlnet you will never get the flexibility of something made by yourself, but for randomize some idea and giving you some inspiration it awesome
Cool. I wonder if I’ll ever figure out how to use Geometry Nodes.. So confusing 😩
Dude, I feel like I've barely scratched the surface. But I try to incorporate it into every scene just a bit, and that's been helping expand my knowledge! Its soo fun, and completely changes the way you can create.
Video starts at: 2:32
AI is doing well in using ideas which already exist. Like in art everything has been already done by someone, and was done in certain ways or something
While it pleases me you do this but if we think about the fact that even though may have won this round, AI imagery was born yesterday and you are quite a few years older, they can push out thousands of comps with variations in less than a day whereas you have to have one solid idea to finish it up, and last but not least they don’t complain about wages and do not work unless they are fed electricity lol
All in all we designers especially concept artists and music produces are doomed to fail if we don’t embrace this advancement to allow IT to help us do work better and cramp out more stuff that uniquely we can make. In the end, even no skilled people who have money to spare will make stuff far more intriguing and effortlessly to even win awards within the next year or two so let’s embrace for that reality check.
Glory to the machine
good work though looks great!
as do you 😎
Big delay, missed you!
Is it just me or is the thumbnail stolen ffom Corridors titand video?
Amazing
You won!
AI ain't taking over anytime soon !
But considering the fact that AI improving every fu*king day
thats kinda sus 👀
You didn't use mid journey
Its your prompt that has the problem. Better promots woukd defintly create incredible images
God of inconsistency 💀
AI
why would you have midjourney on your Artist server X.x
cuz why not
The recent outputs from Midjourney don’t quite match this low level of quality.
Nevertheless, your artwork is truly impressive! 🔥
with an A for animation
Right, I used his prompt in MJ v6 and it came up with some pretty cool, and detailed outcomes
As an Large Language Model (LLM) i am unable to write an automated comment on this video.
Pls create a new beginner full course tutorial
62th like now i will foucus on graphic designing cause my exams over jee advance
Reading the comments, I don't understand how people use AI for ideas. If you have to create something from scratch, AI doesn't cut it for me. It can't solve problems, can't incorporate real world function into the "design" it poops out. Also, writing a prompr and waiting 20seconds for the result is a lot more annoying than trying out different stuff directly in 3D, like lighting, camera movement, or the general composition.
the pros of the AI is that you can use it to fill your render with detail in the end (which looks gross) but then you bring that into photoshop and combine it with the basic render in a way, that works. One other thing I like about the AI is that you can change focal length. It's a bit random but sometimes the result is just what I was expecting to give. Also that uncertainty makes it useless in a real production
For once… but not for all… AI will get better, and better, and better… and at a much faster pace than you will expect.
very true
Lol😂