This guy has a history of making wild statements like this while also claiming to be joking.
He hasn't even played BoTW or Crimson Desert. In fact, he hasn't played a lot of video games.
Whisper the Wolf just kinda standing there because I couldn't think of anything interesting lmao
Finally forced myself to pick up my pen and this is what happened lol
Haven't read the comics yet, but I think Whisper's neat :)
#RoaringLycanArt #FanArt #Sonic #Whisper #DigitalArt
Gel pen sketch of a male and female anthropomorphic wolf. The pair wield round shields and single edged swords.
Morning sketch with more scrimbloized wolves. I thought something was working with the male but the female one is flubbed to me.
Sketched Triss doing the glasses thing.
I friggin love this card design you came up with. lol
A stylized portrait of Llewelyn as the Queen of Spades. He's wearing a black turtleneck, back facing us, wielding two blades in front of a floral background. He's looking back at the camera with a mischievous look on his face as he licks his chops.
The Knife
Lika enjoying herself underwater! 😁
Commission for @/Dragons4lyfe37 on Twitter
Definitely gotta redraw Cole again when I get the chance tbh.
Casual Sesha enjoying some wine.
I also did a festive drawing for @drawstech.bsky.social of his lovely Cleo 🐺♥
Character belongs to @ermwhat404.bsky.social Thanks for being a cool friend, bro! 🙏
I really felt like drawing ermwhat's cute corgi gal, Maddison. LOL
I think they may be tired of the process of art.
I get it, because I've been there. From what I've read tho, they only want to make pretty images. I think that's fine because AI is here to stay but they need to be 100% transparent about their process.
Yes, having alot of output is nice but what's the point when it's not a true representation of your actual work? You're simply developing a habit of relying on these tools to get the job done.
More advanced users with an idea of what they want will use controlnet. Lazier users will type up a prompt and make minor adjustments after the image is made, which they'll give the most control to whatever program they use.
Also what you're suggesting doesn't help with stylistic changes.
That's mostly if the changes make sense. Depending on the model, lighting and color saturation can be wrong or brushstrokes/detail can seem nonsensical. And any mistake you've made won't automatically be fixed after img2img. It's more better to study from actual artists and real life instead of AI.
Posting!
Even taking an image you believe to be "finished" and passing it through AI as a refiner ultimately will not help you as an artist because not only it changes the original style but you don't learn from the process other than how to use AI more efficiently.
He's an example of a person who sketches the foundation of his images but lets controlNET do the rest, which is what most people using AI will do.
And because he does this, he's not actively learning/studying fundamentals. Even now, his problems with anatomy still persists.
Despite having some degree of control AI output, Shad M Brooks went from having his own personal style, to a more generic look because he's relying on AI models.
He has a detailed overview of his process yet it lacks his original style in a lot of areas.
I can see why people would pick up AI and use it as a tool, but at the end of the day it's taking away creative processes and spitting out an amalgamation of ideas.
At best, AI should be used as brainstorming for ideas. Not replacing parts of your drawing process.
Using AI as a crutch can potentially stunt your growth as an artist. That could be said practicing fundamentals improperly but using AI allows you to somewhat bypass problems that you will face normally (mostly with anatomy and color theory). Even then, other issues arise (bad backgrounds).
There are currently thousands of small scale models (Loras) that are being created from the work of many, many artists, all of them without anyone's consent and being offered to people to use. Saying that people shouldn't be afraid of them is downplaying just how big the problem is.
I got a lot to say.
Models with clean data sets wouldn't produce work similar to someone else's unique work unless it's trained on it. There's a reason why there are tons of loras that are trained on an artist and not always embedded within local models. The work produced would be generic.
Testing out a new tablet!
😩Love this cat girl.
Not the most easiest to impress, but she appreciates the stupidity and attempt. Now get out of the way or get the claw.
The miffed Murf
#digitalart #art #oc #characterdesign #catgirls #murf
Maru Ref sheet!
My cute boyo got a new look!
Happy to see you
@duneflower.bsky.social
Sketch of Sesha I did sometime ago.
Fishnets look really good on her. 👀
