Thanks for all the comments. I enjoyed making this one a lot. Quick tip... you can save 20% on a Meshy subscription with promocode "AIANIMATION". Link: - app.meshy.ai?via=aianimation
I did some tests and found it promising. I wasn't able to generate anything that I could use in a final project, but I believe that perhaps in 6 months or 1 year the tool will be able to generate content of sufficient quality for professional use. I will follow the news.
@@mt_gox why would you get joy from this? When everyone is poor because A.I has stolen their jobs...do you think they're going to be spending their money on entertainment produced by A.I? No...crime will just spike and we can go back to a feudalist society.
Those are all suitable quality to make fun little animations for tiktok / shorts with a quick turnaround or an indie game that leverages having TONS of varied assets that would normally take like 6 months to do on your own. Can't wait for game programming logic to be handled by AI.
Not sure about this certain project, but AI for sure will change whole game dev industry. Basically everything can be at least boosted by it... or even replaced. Once AI is advanced enough, manual coding wont be a thing, same goes for modeling, animations etc. One thing that AI wont replace is human imagination.
I could not get a good mesh out of meshy. It would be great if we could do a full 360 degree rotation of images and have it make a much better mesh like with NERFS.
This is a great channel - consider making deeper tutorial videos on a paid tutorial platform. Your workflows are really great & I can't believe how many discoveries I picked up from 2 of your videos.
That Cartoon Line-Art skull looks fantastic! I've always been a big fan of that "Borderlands" look-- this is fairly close! Always excited to see what Meshy is up to and how creative people are using it! Great video!
I was listening to commentary on Generative AI and the point being made is that if people are looking to make money there potential best bet is for manual labor or mechanical jobs. Like being an hvac repair man or a helicopter /aircraft mechanic. Its just sad. The narrative had always been sold to us as AI will take the jobs that no one wants to do and free us up to do more art. Yes, it takes very little time to generate these models, but it's now been commoditized. Or at least the bottom part of the market will be. And more and more of the creative market will be subsumed as the technology gets better. When you automate things, you also separate a part of yourself from it. Denuding humanity's relationship to what draws us to art in the first place. It's a complicated issue, but generative ai doesn't make me feel good as a creative, personally.
@@Ash-nh6li I think there will always be a market for stuff made by humans. As someone who works in the media landscape, I'm concerned the percentage of people who will be able to make this work as a means of providing income is a fraction of what we've known it to be. I know folks are working on this, but how will we know what's made by humans? After a lot of negative rumination around the topic, I've finally decided to continue my path as a creator. Not by using Gen AI tools, but by stepping up the complexity of what I aim to do. I poke my head in here to see how these things might be helpful to some future creative process, like AI assisted rotoscoping is now. So far though, all I identify these things with is labor REPLACING tech, not labor enhancing tech. I humbly accept this could be because of a lack of creativity or understanding on my part. In general though, most of this stuff continues to be deflating and I think we've yet to see the full extent of the fallout on society.
@@wellheytherehihowyadoin I'm using AI everyday now for my 3d job. What I can do now in one week would have taken me months to do without it. It's only going to get faster and yes there will be a large amount of current workflows replaced. I think we are just going to get more involved with the craft of world building in the future. AI is still pretty aimless as impressive as it is, there will still need to be a human creative to direct the AI. I am holding out for the positives and in no time we are going to see some very complex masterpieces and a TON of really simple poor thought out projects by the masses.
Thanks. Yeah I'm going to revisit it soon to explore the image to 3D approach. I know they had an update in the works, as the existing version was'nt the best. Plus with Stable Diffusion, recently releasing their 3D-model video out, which does a good job of turning an image into a video that spins aorund a 'model' as a 360 video. Which could be used to help generate a 3D mesh (via a few approaches) so perhaps they and others will integrate something like that to improve the image to 3D model approach... We'll see.
This is a great tutorial, thank you for looking at 3d generation and also incorporating it into After Effects. I can't help thinking that the Vision Pro has just made 3d skills even more important
Thanks. Yeah absolutely, I think if/when millions of people are using VR/AR on a daily basis there's going to be much more demand for 3D content. (*I've watched too many reviews/unboxings of the Vision Pro 🙂).
@@AIAnimationStudio Cheers man. Been following the AI stuff for a while, first with Disco Diffusion, then Stable... but kind of haven't kept up - so good to watch your videos. Keep up the good work man!
Even if it's still relatively primitive, it still would be useful to use with the rise of "PS1 Graphic" games coming out. This could be a style of its own. AI 1.0 style. But yeah, hope it does get better in the end, and hope AI will cover the more tedious areas of 3D. Creating the model seemed like the most fun to me, I don't want AI to substitute that. But texturing, rigging, animating, UV Mapping, programming, coding etc. Those are more the areas I would like AI to cover, right now AI is mostly taking over the "fun" aspects of creating 3D models and games.
@@IagoRodriguezDopico AI lacks individuality, taste and the ability to produce anything new. AI will not be able to sew UVs, build quality topology and have the same artistic touch as a human does. It will never be good and the slop will stand out.
Excellent 3D design tutorial for AI animation! Everything you need to know to use AI technology to produce stunning models and textures is covered in the video.
It can be useful for generating quick background and out of focus stuff in a scene, but please guys do proper modeling and animation for the main subject..
Damn paywalls ... I'm not sure which tier I was on when making this... but yep, like all these AI platforms they keep evolving as they figure out how to keep the lights on.
thx a lot for your awesome videos ! When making a GLB file using Zoe Depth, i have a very poor/blurry quality texture on the 3D asset even at full resolution in After FX. Do you know a way to improve this ? Thank you.
There's definitely some loss to the quality to the texture. As it's applied as colours on the mesh (I forget the name for it right now) rather than as a texture map material. (So your original image used by ZoeDepth isn't actually used as the texture itself). So, unfortunately, remaining pretty zoomed out is the best option. It's a limitation with ZoeDepth as it currently stands.
@@AIAnimationStudio Thank you for your answer ! Also you just solved a mystery for me, i was trying somehow to edit it with a GLB Editor but there was no texture .... because thx to you i now know it's just colors ! You just saved me a lot of time, thx a lot ! i wish it can be improved someday because it's just a better to make small 3D camera moves than Displacement & depth maps. Let me know if you find a killer tip about this in the future ! Thx again !
@@landdweller7523 because the topology and UV maps are absolutely atrocious in terms of how they're configured and are not ideal at all for bone driven deformation.
I just used this tool to make a chess pawn, and... it did not look at all light a pawn sadly. I really wanted to like it -- we needed this years ago -- but most 3D modeling AI tools are useful enough yet to replace modeling by humans.
Certainly not optimised for rigging, yet. Hopefully that’s something will see in meshy and similar products as the quality of models reaches the quality that more users will be looking to rig. I could see that almost being a separate tool that will be in demand too if it could be applied to existing models and refine the mesh at key areas (I.e. optimising shoulder and elbow joints has been a pain for everrrr)
What does the topology actually look like though? Every one of these I've tried creates trash geometry that would take an artist almost as much time to clean up as to make from scratch. Plus I would prefer to be able to give it actual images to generate from rather than text.
I think its great for people making their own video games, side projects, etc. But I'm not sure I see this as anything other than possibly a tool? For blocking out? I guess? For my job I need to make specific aircraft parts that have no descriptor, I have no idea what the name of it is, nowhere to look that information up, and I only get references provided to me as I cant go up to the plane and take it a part obviously. So this isn't useful to me personally in my job but I can see how maybe its useful to someone who does renders? doesnt care about topology, doesn't need something super specific down the size, etc.
Thanks for posting, that's a really interesting role you have and yeah I can't think how Meshy (as it currently stands) can help with that. Perhaps once the image to model generation improves via Meshy or another platform there might be some scope, but I'd assume for your use case you'd need 100% exact control so manual modelling will always be needed. But perhaps image to models can help give you a reference model that's some what useful.
@@graphicdestination968 . Not possible yet. But found a nice work around for now in this video th-cam.com/video/7aYy9H7jQuQ/w-d-xo.htmlsi=hTQqArFeI9SbSVWo. Plus some other tools that might help (with a bit more exploration) here th-cam.com/video/2dsxC6GnJlo/w-d-xo.htmlsi=8ocWIQY3YSrPnUb7
You dont get exactly what you want, I wish there were more tools to finetune the generation process. Right now it's almost like generating and hoping for the best. BUT I do see the potential in the future for this.. ASwell if you need background caharcters or art, it can be really good.
Fixing ai mistakes cost more time then creating from scratch, that’s what I feel all the time when making images from midjourney takes ages to fix the errors in photoshop, I was already super worry about my job but these ai tools making so many mistakes not sure if they will get there to make them without these errors, I many times create a image from ai looks great then looking to the image see one error then the next and next and next in the end I fixing 2 hours of errors. Terrible experience, I’m not 3d expert but I think here will be the same bs, sure looks great but first you don’t get what you wanted and second fixing all these and then you still have all the faces of the model bad state
I’m pretty much 100% sure that you can’t use any of these models for real work since they all broken, sure maybe for some background elements this could be ok
Incredible. I'm a little surprised the paid membership includes ownership of the assets generated. Isn't data generated from a blend of assets owned from other sources? Excuse my naiveness.
It's my understanding that AI doesn't use other content directly, rather uses a combination of them to create its own, unique piece. Pretty much what a normal artist does but in a mili-fraction (if that's a word) of the time. So technically, they're not stealing anyone else's work any more than an artist who uses say, marvel comics as inspiration is stealing another artists work.
@@Eric-cf8lp Not saying I agree with it, I hate AI art, it's just that's how it works under the hood and why they can get away with essentially stealing other artists work while getting around the law.
Let's face it. the quality is not good for now. but wait 10 to 20 years! Just like the video games. compare them from the 90s up to now. look at the improvement. let's embrace AI :)
@@sasnad3 So what's your talent? Typing words? If you type words into the prompt boxes like you type on TH-cam i doubt that we, real 3D artists, have anything to fear. Have a good one :D
@@TheGIRLIoana ok first you need to tell me why you assumed that I'm not a 3d artist? Secondly i just stopped u from giving up and crying about it, because that's lame. If you are a creative artist you will do good anyway. I make thousands of $ selling 3d models. And i know tht people will opt for a proper model that works across many platforms and also because they are reasonably priced and work with any game engine hassle free. Even if Ai starts making really good stuff, which I doubt..still il be selling.
@@TheGIRLIoana the moment u accept ur fate, you do nothing to change it. And that's a problem that needs to be addressed. Apologies if you didn't like my criticism.
Good question. There are quite a few areas of animation/design I've not seen covered by AI yet. I'm sure it's in the works by various companies. Lots of problems to solve but I suspect it's coming.
Good Morning, brother. I would like to talk to you about business and commerce here in Brazil with your teaching. How can I have a private conversation with you?
This will be a critic for the ai technology and not for the video.(The video vas understandable) To start of: before i can even think of an use of this i wanna know how it deals whit topology (this is in my eyes the most important, if topology is bad the the other ones will 100% be bad), normal maps, textures, riging(how many bones it gives, how it places them, and last but not least weight painting). If it gives anything close to unfinished sculpt whit bones, random weights, and randomly cut maps, then it will be hard as f to edit it and clean it up afterwards. (Also just for demonstration thet bear is not really well riged, one look at its arms on 8:24 and any given people who has played any videogames or seen any 3d animated movie can tell thet something is really off, this is probably topology failure and because of thet it can not be well weight painted and animated) If i woud have to use it right now i woud maximum only use the animation feature (it woud probably need some clean up), but whit a self-made and rigged 3d model, or just a realy based out starting point for modelling. And just to be clear i am not saying thet its bad, i am saying thet it has a looooooooong way before it it usable any area of 3d modelling. But its an acceptable start for ai in 3d modelling, i believe it will never replace 3d artists because it can only combinate and can not create anything new, it will always need more training data or it will put out the same results after some time, but it will make the process of 3d modelling a lot easier. Hope this review helps.
Thanks for sharing. Agree on all counts. Its impressive, and as with all the things will only get better... but to be used for anything professional right now the use cases are pretty limited. The new texturing improvements added since posting this video are pretty cool. I'll try and cover that soon.
Thanks for all the comments. I enjoyed making this one a lot. Quick tip... you can save 20% on a Meshy subscription with promocode "AIANIMATION".
Link: - app.meshy.ai?via=aianimation
I did some tests and found it promising. I wasn't able to generate anything that I could use in a final project, but I believe that perhaps in 6 months or 1 year the tool will be able to generate content of sufficient quality for professional use. I will follow the news.
In 1 year it will die like NFT
@@Lyco0n in 1 year you guys will be working at mcdonalds after being replaced by ai 🙂
@@mt_gox remind me in 1 year when I still work remotely from anywhere
@@mt_gox why would you get joy from this? When everyone is poor because A.I has stolen their jobs...do you think they're going to be spending their money on entertainment produced by A.I? No...crime will just spike and we can go back to a feudalist society.
Maybe more time will be needed, but I still see potential in Meshy nonetheless.
I used Luma AI Genie to do this, which is free, but this has more features. These are great for assets.
Now I'm starting to get interested. Although it's reminiscent of early 3D game assets, it has really taken leaps and bounds!
Those are all suitable quality to make fun little animations for tiktok / shorts with a quick turnaround or an indie game that leverages having TONS of varied assets that would normally take like 6 months to do on your own. Can't wait for game programming logic to be handled by AI.
Try Luma Genie.
can you wait for AI to do absolutely everything for you? like... what ever will be left for humans to do? to aspire to... this is just... sad
Not sure about this certain project, but AI for sure will change whole game dev industry. Basically everything can be at least boosted by it... or even replaced. Once AI is advanced enough, manual coding wont be a thing, same goes for modeling, animations etc. One thing that AI wont replace is human imagination.
I could not get a good mesh out of meshy. It would be great if we could do a full 360 degree rotation of images and have it make a much better mesh like with NERFS.
This is a great channel - consider making deeper tutorial videos on a paid tutorial platform. Your workflows are really great & I can't believe how many discoveries I picked up from 2 of your videos.
To think that I only subscribed two days ago and you give me this gem! Thanks man.
Awww... great!... glad you liked it. Plus thanks for the sub.
bro is proud to get a reply from creators
Wow this tool is really impressive.
That Cartoon Line-Art skull looks fantastic! I've always been a big fan of that "Borderlands" look-- this is fairly close! Always excited to see what Meshy is up to and how creative people are using it! Great video!
Thanks Tim. Yeah, I’m tempted to try and generate a full scene in that style.
So cool that they show community prompts like text to image sites.
Very impressive! Things are evolving so quickly. Thank you for this! ❤
Cheers. Yeah it's moving quickly and seems like these AI 3D gen tools are now creating really useful content to compliment more traditional 3D work.
I was listening to commentary on Generative AI and the point being made is that if people are looking to make money there potential best bet is for manual labor or mechanical jobs. Like being an hvac repair man or a helicopter /aircraft mechanic.
Its just sad. The narrative had always been sold to us as AI will take the jobs that no one wants to do and free us up to do more art.
Yes, it takes very little time to generate these models, but it's now been commoditized. Or at least the bottom part of the market will be. And more and more of the creative market will be subsumed as the technology gets better.
When you automate things, you also separate a part of yourself from it. Denuding humanity's relationship to what draws us to art in the first place.
It's a complicated issue, but generative ai doesn't make me feel good as a creative, personally.
maybe there will be a market for human made things, including games
@@Ash-nh6li I think there will always be a market for stuff made by humans. As someone who works in the media landscape, I'm concerned the percentage of people who will be able to make this work as a means of providing income is a fraction of what we've known it to be. I know folks are working on this, but how will we know what's made by humans?
After a lot of negative rumination around the topic, I've finally decided to continue my path as a creator.
Not by using Gen AI tools, but by stepping up the complexity of what I aim to do. I poke my head in here to see how these things might be helpful to some future creative process, like AI assisted rotoscoping is now. So far though, all I identify these things with is labor REPLACING tech, not labor enhancing tech. I humbly accept this could be because of a lack of creativity or understanding on my part. In general though, most of this stuff continues to be deflating and I think we've yet to see the full extent of the fallout on society.
@@wellheytherehihowyadoin I'm using AI everyday now for my 3d job. What I can do now in one week would have taken me months to do without it. It's only going to get faster and yes there will be a large amount of current workflows replaced. I think we are just going to get more involved with the craft of world building in the future. AI is still pretty aimless as impressive as it is, there will still need to be a human creative to direct the AI. I am holding out for the positives and in no time we are going to see some very complex masterpieces and a TON of really simple poor thought out projects by the masses.
Thanks I'll give it a go for my project 😮
Very nice!
Thanks!
Amazing
really cool video. Quite interesting to see what is possible, thanks for sharing.
This is Crazy Useful
Amazing!! thank you , it would be really great to see videos for generating 3d models from images ,thanks
Thanks. Yeah I'm going to revisit it soon to explore the image to 3D approach. I know they had an update in the works, as the existing version was'nt the best. Plus with Stable Diffusion, recently releasing their 3D-model video out, which does a good job of turning an image into a video that spins aorund a 'model' as a 360 video. Which could be used to help generate a 3D mesh (via a few approaches) so perhaps they and others will integrate something like that to improve the image to 3D model approach... We'll see.
They should have a NN model for all these animation in mixamo. That way, you could integrate that same animation into a NN text-to-video workflow.
Very efficient video, thank you !
This is a great tutorial, thank you for looking at 3d generation and also incorporating it into After Effects. I can't help thinking that the Vision Pro has just made 3d skills even more important
Thanks. Yeah absolutely, I think if/when millions of people are using VR/AR on a daily basis there's going to be much more demand for 3D content. (*I've watched too many reviews/unboxings of the Vision Pro 🙂).
Meshy is very good!
That is so cool man!! Thank you🙏
This is awesome! People who claim "AI is the death of creativity" never actually cared about artists.
Yay exciting
They all look like WoW meshes and textures.
I’m just learning 3D and now IA can make it faster and better 🥲
Honestly, I think 3D skills are going to be so useful for many, many years to come.
Brilliant tutorial and very informative ❤
Thanks very much :-)
I'm curious to see what they'll look like after I poison this LLM with Nighthsade.
Based
Hello John!! Hope you're doing well mate, nice videos very ueful :)
Cheers Olly. It's been ages!... love the work I've seen you tackling over the years.
@@AIAnimationStudio Cheers man. Been following the AI stuff for a while, first with Disco Diffusion, then Stable... but kind of haven't kept up - so good to watch your videos. Keep up the good work man!
Even if it's still relatively primitive, it still would be useful to use with the rise of "PS1 Graphic" games coming out. This could be a style of its own. AI 1.0 style.
But yeah, hope it does get better in the end, and hope AI will cover the more tedious areas of 3D. Creating the model seemed like the most fun to me, I don't want AI to substitute that. But texturing, rigging, animating, UV Mapping, programming, coding etc. Those are more the areas I would like AI to cover, right now AI is mostly taking over the "fun" aspects of creating 3D models and games.
Cool!!
Genio, excelente tutorial
Wow....Amazing...😊
Great tips!
well done easy to follow along
Very informative 🎉
Wow! Amazing video🎉 thank you👍
Awesome. Pleased you liked it.
Meshy AI + reallusion iclone workflow tutorials for better rigging plz
@@graphicdestination968 . One day perhaps
🎥🔥🔥🔥🔥needed this…thank you
Valuable stuff, Thank you.. Expecting more on the 3d ai tools.
A si no lo tienes que estudiar jajaja jajaja
Nice info, Hi, I am Ravi From India, Thankyou
Thanks. Hi from the UK.
Awsome❤
All those 5 years of me learning Blender , Puff
haha man, they all look shit, keep the old way please
Itll still have to be polished up, so keep learning!
the models are all shit, lol how normies freak out over this
Well, they freak out because in no time they will be great. This is the sea receding before the tsunami
@@IagoRodriguezDopico AI lacks individuality, taste and the ability to produce anything new. AI will not be able to sew UVs, build quality topology and have the same artistic touch as a human does. It will never be good and the slop will stand out.
Excellent 3D design tutorial for AI animation! Everything you need to know to use AI technology to produce stunning models and textures is covered in the video.
You sir are a diamond
💎 .. Cheers! 👍
Thanks for putting 3D artists out of a job based SOLELY on input from 3D artists. Well done!
He didnt make the app. lol
Promoting it is contributing to its success, use your head@@Darksagan
very helpful cheers
Excellent.
amazing
Very cool
very nice thank you very much sir.
It can be useful for generating quick background and out of focus stuff in a scene, but please guys do proper modeling and animation for the main subject..
Seems like he has PRO to get all these extra features. Oh paywalls...
Damn paywalls ... I'm not sure which tier I was on when making this... but yep, like all these AI platforms they keep evolving as they figure out how to keep the lights on.
thank´s its cool
Epic
Awesome content. How can I turn my PNG Logo into a 3D asset. is there a tool for that? Mine has 3 color variants. Thanks
Have you tried upscaling the frames of the video with Krea?
They also offer a unity plugin.
Will be usable for pro in 1 year
thx a lot for your awesome videos ! When making a GLB file using Zoe Depth, i have a very poor/blurry quality texture on the 3D asset even at full resolution in After FX. Do you know a way to improve this ? Thank you.
There's definitely some loss to the quality to the texture.
As it's applied as colours on the mesh (I forget the name for it right now) rather than as a texture map material. (So your original image used by ZoeDepth isn't actually used as the texture itself).
So, unfortunately, remaining pretty zoomed out is the best option. It's a limitation with ZoeDepth as it currently stands.
@@AIAnimationStudio Thank you for your answer ! Also you just solved a mystery for me, i was trying somehow to edit it with a GLB Editor but there was no texture .... because thx to you i now know it's just colors ! You just saved me a lot of time, thx a lot ! i wish it can be improved someday because it's just a better to make small 3D camera moves than Displacement & depth maps. Let me know if you find a killer tip about this in the future ! Thx again !
Can you use ai to create 3d prints?
Why is iy the picture looks great but when I erxport a .stl fill all the detail is lost?
If you're just messing around sure but if you're looking for something to generate objects that can be used in games you can skip this.
Could you tell me why?
@@landdweller7523 because the topology and UV maps are absolutely atrocious in terms of how they're configured and are not ideal at all for bone driven deformation.
Thanks for your tutorials, very helpful for me!
I just used this tool to make a chess pawn, and... it did not look at all light a pawn sadly. I really wanted to like it -- we needed this years ago -- but most 3D modeling AI tools are useful enough yet to replace modeling by humans.
Does that bear not need retopology , how it is optimized for rigging
Certainly not optimised for rigging, yet. Hopefully that’s something will see in meshy and similar products as the quality of models reaches the quality that more users will be looking to rig. I could see that almost being a separate tool that will be in demand too if it could be applied to existing models and refine the mesh at key areas (I.e. optimising shoulder and elbow joints has been a pain for everrrr)
What does the topology actually look like though? Every one of these I've tried creates trash geometry that would take an artist almost as much time to clean up as to make from scratch. Plus I would prefer to be able to give it actual images to generate from rather than text.
I think its great for people making their own video games, side projects, etc. But I'm not sure I see this as anything other than possibly a tool? For blocking out? I guess? For my job I need to make specific aircraft parts that have no descriptor, I have no idea what the name of it is, nowhere to look that information up, and I only get references provided to me as I cant go up to the plane and take it a part obviously. So this isn't useful to me personally in my job but I can see how maybe its useful to someone who does renders? doesnt care about topology, doesn't need something super specific down the size, etc.
Thanks for posting, that's a really interesting role you have and yeah I can't think how Meshy (as it currently stands) can help with that.
Perhaps once the image to model generation improves via Meshy or another platform there might be some scope, but I'd assume for your use case you'd need 100% exact control so manual modelling will always be needed. But perhaps image to models can help give you a reference model that's some what useful.
Can I use this 3D model which is generated in here as commercial purpose such as game?
How would we generate 3D models for parts (like body parts) and then combine them?
So then we can import into a slicer and add auto supports and print!
also how to do facial animation which are created in meshy AI?
@@graphicdestination968 . Not possible yet. But found a nice work around for now in this video th-cam.com/video/7aYy9H7jQuQ/w-d-xo.htmlsi=hTQqArFeI9SbSVWo.
Plus some other tools that might help (with a bit more exploration) here th-cam.com/video/2dsxC6GnJlo/w-d-xo.htmlsi=8ocWIQY3YSrPnUb7
ty
You dont get exactly what you want, I wish there were more tools to finetune the generation process. Right now it's almost like generating and hoping for the best. BUT I do see the potential in the future for this.. ASwell if you need background caharcters or art, it can be really good.
Can you create modular characters
Can this be Exported to other Apps and edited?
Yeah, you can download the model in various 3D formats to edit in your preferred 3D package. i.e. Blender, Cinema 4D, Maya etc.
Fixing ai mistakes cost more time then creating from scratch, that’s what I feel all the time when making images from midjourney takes ages to fix the errors in photoshop, I was already super worry about my job but these ai tools making so many mistakes not sure if they will get there to make them without these errors, I many times create a image from ai looks great then looking to the image see one error then the next and next and next in the end I fixing 2 hours of errors. Terrible experience, I’m not 3d expert but I think here will be the same bs, sure looks great but first you don’t get what you wanted and second fixing all these and then you still have all the faces of the model bad state
I’m pretty much 100% sure that you can’t use any of these models for real work since they all broken, sure maybe for some background elements this could be ok
absolutely loved it ❤️
He's really good..... at spending money.
Is this copyright free?
something like that that I wanted to do but with open source, do you know any alternative that I could have locally?
I've seen some complex workflows with people generating 3D things via Comfy, but not as downloaded 3D files.
Incredible. I'm a little surprised the paid membership includes ownership of the assets generated. Isn't data generated from a blend of assets owned from other sources? Excuse my naiveness.
It's my understanding that AI doesn't use other content directly, rather uses a combination of them to create its own, unique piece. Pretty much what a normal artist does but in a mili-fraction (if that's a word) of the time. So technically, they're not stealing anyone else's work any more than an artist who uses say, marvel comics as inspiration is stealing another artists work.
@@skybluerob1 wow, what a naive take.
@@Eric-cf8lp Not saying I agree with it, I hate AI art, it's just that's how it works under the hood and why they can get away with essentially stealing other artists work while getting around the law.
Does it do img 2 3D? If not, whats the best way to do that?
It does have image to 3d but it’s not (yet) as good as their text to 3D option.
Hi, I am in need of someone to create a model for me using blender and I am willing to pay for the work. Would you be able to help me with this?
What kind of model do u have.
i cant` import the files in after effects 2024. how can i do that?
Let's face it. the quality is not good for now. but wait 10 to 20 years! Just like the video games. compare them from the 90s up to now. look at the improvement. let's embrace AI :)
yeah! let's accept the fact that us, real 3D artists will get replaced! Awesome! Can't wait! :)
@@TheGIRLIoanayou need to go into a corner and cry. Lame
@@sasnad3 So what's your talent? Typing words? If you type words into the prompt boxes like you type on TH-cam i doubt that we, real 3D artists, have anything to fear. Have a good one :D
@@TheGIRLIoana ok first you need to tell me why you assumed that I'm not a 3d artist? Secondly i just stopped u from giving up and crying about it, because that's lame. If you are a creative artist you will do good anyway. I make thousands of $ selling 3d models. And i know tht people will opt for a proper model that works across many platforms and also because they are reasonably priced and work with any game engine hassle free. Even if Ai starts making really good stuff, which I doubt..still il be selling.
@@TheGIRLIoana the moment u accept ur fate, you do nothing to change it. And that's a problem that needs to be addressed. Apologies if you didn't like my criticism.
the models ARE BAD IN QUALITY AND THE TRUTH FOR CREDITS IS NOT COST EFFECTIVE. STOP ENCOURAGING PEOPLE TO SPEND ON VERY BAD QUALITY MODELS.
I just used it and it had major texture issues is there a way to fix it
They've rolled out a way to select areas of the model and redo and improve texturing. Yet to try it out myself though.
What about animated engineering drawings? Why does AI never ever talk about engineering applications?
Good question. There are quite a few areas of animation/design I've not seen covered by AI yet. I'm sure it's in the works by various companies. Lots of problems to solve but I suspect it's coming.
What are 3d?
Can I save em as stl file?
Yes they have stl as one of the download options. 👍
Good Morning, brother. I would like to talk to you about business and commerce here in Brazil with your teaching. How can I have a private conversation with you?
Hey, you can get in touch via contact details at aianimation.com/contact
The Work is make easy but artistic is lost
This will be a critic for the ai technology and not for the video.(The video vas understandable)
To start of: before i can even think of an use of this i wanna know how it deals whit topology (this is in my eyes the most important, if topology is bad the the other ones will 100% be bad), normal maps, textures, riging(how many bones it gives, how it places them, and last but not least weight painting). If it gives anything close to unfinished sculpt whit bones, random weights, and randomly cut maps, then it will be hard as f to edit it and clean it up afterwards.
(Also just for demonstration thet bear is not really well riged, one look at its arms on 8:24 and any given people who has played any videogames or seen any 3d animated movie can tell thet something is really off, this is probably topology failure and because of thet it can not be well weight painted and animated)
If i woud have to use it right now i woud maximum only use the animation feature (it woud probably need some clean up), but whit a self-made and rigged 3d model, or just a realy based out starting point for modelling.
And just to be clear i am not saying thet its bad, i am saying thet it has a looooooooong way before it it usable any area of 3d modelling.
But its an acceptable start for ai in 3d modelling, i believe it will never replace 3d artists because it can only combinate and can not create anything new, it will always need more training data or it will put out the same results after some time, but it will make the process of 3d modelling a lot easier.
Hope this review helps.
Thanks for sharing. Agree on all counts. Its impressive, and as with all the things will only get better... but to be used for anything professional right now the use cases are pretty limited. The new texturing improvements added since posting this video are pretty cool. I'll try and cover that soon.
interesting neuron, the only bad thing is that some models come out crooked
And you don't show a single topology ????
This is extremely good news