AI can easily reproduce that. The question will be how feasible the technology is. Additive manufacturing/3D printing in stone or word is undoubtedly possibly but no doubt an absolute pain in the ass to implement so nobody will ever own one.
You mean wood, right? What about metal, are they harder than the two other materials?
Of course, phoneposting is a pain in the ass sometimes. Apparently 3D printed wood exists but is finicky with the printer and it's simply "wood filament" and you get some knockoff of the actual grain on rare woods and none of the hardness, weight, etc. 3D printed stone exists too but is the same shit--you get god knows what sort of rock and the 3D printer can make it look like marble or whatever but without the weight, hardness, durability, etc. Tech doesn't seem there yet to actually print a mahogany statuette or a marble sculpture.
That already exists and has a lot more obvious paths to improvement since it's just fucking around with ceramic powders or even just a glorified sort of metal casting like people have done for millennia.
I swear I saw someone use a 3d printer to create a cast. I mean obviously bronze or similar so it's pretty fucky and that sort of sand casting will never have the same 'resolution' as a 3d printer. Idk much about that side of manufacturing but I'm fairly sure the cost of like creating the design of the hole is more expensive than making the hole. As in the majority of the money in a company like games workshop is spent on the artists/designers rather than creating the actual metal things the plastic goes into. If you could get a computer to do that labour it'd probably cut down the costs pretty heavily. Or if you only want to make a few then you could probably pretty easily get a 3d printed mold and cast whatever with resin or similar and have a pretty decent resolution. Hopefully that doesn't happen too soon, I'm already looking at buying a new gpu to fuck around with AI better; don't want to have to go buy a resin printer at the same time too.
The fact that there has been almost no development on ML plugins for something like Blender or Maya that can auto retopo a mesh from something like Zbrush or make actual good UV maps is fucking bullshit. This technology has been around forever in the scale of things and yet nothing. Procedural 3D gen models like TRELLIS are really only useful as tech demos.
I've had some luck with Meshy for image to 3d model generation, but it can't be too high detailed or disjointed if you want to port over the textures too. If you are doing blobby things and robots and people it's a fine tool. But I'm just importing models into TTS and I've cleaned one up to print and it's wonky. I saw a pretty good photogramaphy (Sic) ai tool online but I can't find it anymore and didn't have the tools to use it at the time.
There's the vegoilguy and Robinson foundry who are both overly thorough in there mold making.Robinson less so cuz he's a chad. Vegoilguy is a drain on the electrical grid but his bowling alley animations make up for it.
Before AI art got better, one could give a description and maybe share a reference pic with a human artist for an often overpriced commission. Such a commission -- if it is even available -- can take quite awhile to make, and that human artist can be rather difficult to deal with. Maybe the human artist could even start BS drama. And of course human artists -- at least Western ones -- can be irrationally fanatical on copyright and treat the slightest copying as a horrible "art theft"... maybe even going to court.
You could take it very far indeed given how easy it is to find examples of people randomly accusing actual human made art, mediocre as it is, as being AI generated.
While artists may claim AI art has "mistakes no human would ever make", you can find similar errors in art made by natural intelligence: bad hands, warped backgrounds, etc.
Yes, exactly, and the tards saying it's an ai tell are both behind the times and just incorrect. Anybody who's not a skilled artist will fuck up on hands and plenty of other details, maybe not in a way an earlier AI image gen did, but try asking asking most of the xitter artists pushing this meme argument to draw you a realistic person or even an anime character with 5 fingers.
A big fat example of that is "Sonichu" comics -- though "Honor Roll" has "exceptionally" low skill combined with a freakish pen grip.
Oh yea people already do that. But in general it's a bit different to coding. You don't code for fun, you likely don't really have a passion for coding. For the end product maybe, but the actual task of coding is just a means to an end. For art it's the opposite; people paint as a form of recreation. If you tell them they can automate half their workload they don't hear that as 'you can make twice the number of images' but as 'you can cut the amount of what you enjoy in half'. Most of the mistakes AI does can be fixed in less time than it takes to generate the image. I've had several images that come out almost perfect aside from a logo is fucked up or there's an extra finger. It takes longer to open gimp than it does to remove the extra finger and replace the logo with what you want. Just stick that back into the AI as a base image and you're done. But realistically artists just don't want to do that and because they're just posting shit on twitter they have no real reason to increase output; they don't have a boss demanding they draw quicker, unless you want to count algorithmic shit but even then that's not getting paid.
Bad hands are a problem for many mediocre artists, who will go out of their way to hide hands out of frame because they can't draw them.
Yaoi hands have been a meme for longer than most of the people complaining about AI have been alive too
Incorrect. Coming up with a cool solution to a problem or designing architecture or logistics is awesome. I've used AI to make Liquibase/Kubernetes files, and repos. Basic shit like database ORM entities, and unit tests is what it's best at. Know how to fix those and you have a solid foundation on which to build an application.
I still sort of hope the threat of AI will improve some artists' shitty work morale and communication. I meet on daily basis a really good one, who whines about AI constantly and also not having money, but when asked to do commissions by people I recommended her to, she doesn't even respond at all because she ''did not feel like it'' , only to those who are worthy of response according to some criteria I am not aware of, amount of money is not one. So her potential customers go for AI slop (other choices were too busy with school or transitioning at the moment). Some people just deserve to lose their livelihood to AI. Of course I don't say this is a common thing, but some artists' unreliability and mental illnesses play a role and worsen the situation of the responsible ones too.
A number seriously think all art should cost money, and it's "entitled" to think otherwise. And then they wonder why they lose to AI.
