Freelancer604
patreon
March Previews and Roadmap
🕑 Added 2023-03-06 00:30:09 +0000 UTC**In-post image links on patreon appear to be broken at the moment. I will upload them to the slideshow above.
Wow its March already, one more month, and I will finally be out of this cold Canadian winter! For this month, the pose packs that I am preparing for you will be:
- Awpzzggg's Dance Poses 02
- Portraits & Selfies 01
For the first pack, you can expect a hefty set of dynamic dance poses that you may recognize from some iconic Kpop MVs. The second pack will be stuffed full of cute selfie poses that are meant for close-up shots. I have received a lot of requests for selfie poses over the last year, but to be honest I find it daunting because these poses tend to be nuanced and involve hands, which are difficult to get right. I will finally attempt them this month.
I have not decided on an idea for the exclusive pose of the month at the time of writing, but I can say that I am itching to make something NSFW themed, given the relatively tame roadmap for this month.
Character of the month:
This month's character "Grace," seen above, will have a western aesthetic (or at least western in the way that JRPGs depict westerner girls). Grace will take inspiration from the "kind hearted priestess healer girl" archetype that is in basically every JRPG ever, and come in an outfit to match. For her timed exclusive alter-ego card.. I am thinking of making an evil/witchy version of her... I will add more pictures of her as I get closer to completion. I am currently torn between the platinum blonde and silver hair look... I may just end up releasing both.
I do not have many pictures of Grace yet, but I will update them here as they come.
New Usecase for Poses. Regarding Stable Diffusion Control Net:
An exciting new feature was added to Stable Diffusion recently called control net. It enables the AI to copy the subject's pose from one image and apply it to its own outputs. This is significant for us as Illusion Studio players, because it means that our poses and scenes are now directly transferrable to stable diffusion. All you need to do is to take a studio picture of a character with a pose from an angle that you like. Then use that as part of your input via the controlnet plugin, as demonstrated here:
March AI albums Preview:
For this month, I have released 6 AI photo albums of various sizes. I did a LOT of experimentation with Stable diffusion's various custom models and features to try and get to a point where I am ready to perform commissions for the new Partner patron tier.
Here are some preview images from the new albums:
Unrelated:
There has been a lot of heated discussion in the art sphere about AI-generated art and its disruptive nature. The crux of the issue is that AI has learned to create art at the level that it does today by stealing from and trampling over the intellectual property rights of human artists. It has done so without consent or compensation to human artists.
You didn't ask, but here is my personal take on the issue. I think that the battle is already lost. We are at the beginning of a revolution that will hurt many artists in the short run, but will be greatly beneficial to humanity as a whole. Expectations will need to change. Artists who make a quick buck on Etsy for quick arts will likely see their market shrink dramatically, as anyone will be able to quickly generate their own for their blog or small business using AI. Like with farming during the industrial revolution, the number of people working in the field will decrease, and the quality/quantity of output will increase. I think that a lot of artists fighting this losing battle against proliferation of AI art will soon see that no matter how good AI art gets, people will always be attracted to hand-made bespoke things, and that so long as they have passion and style, their work will always be valued, perhaps even more than before as the competition dwindles.
I would also like to add an extra spicy opinion: As crucial as human art was in the first phase of AI learning, our human contributions to AI's ability to do art will soon be dwarfed and made insignificant by AI learning and iterating on its own art. As people generate more images, they will select for the best outputs and post them online (as I do). This will create an accelerated process of natural selection. The desirable images will feed back into the data sets that the AI learns from, ever improving the AI's ability to cater to our tastes. Feel free to add your thoughts to this discussion below in the comments.
As for our community here, I think its a universal win. We are just here to celebrate and behold beauty, and there will be plenty of that coming our way. Rejoice, friends! 🥰
In other unrelated news, have you seen Wo Long: Fallen Dynasty? That game looks sick and I'm really hoping that I will find the time to play it this month. That studio's previous game, Nioh 2 was one of my favs of all time, combining the best parts of Darksouls-like bosses and Diablo-esque itemization.
Comments
Alessandro
Do you also plan to write very basic tutorials on AI generation in the future? That would be lovely as well :)
