SkyCogs's avatar

SkyCogs

Exploring dreams & dimensions
4.6K
Watchers
332 Deviations
20.1K
Pageviews
CrunchiestKibble
Otaganeki
answ234689
justanadventurer
Supermochi1
ashvillia
Wolfgang-Caesar
willinvadesearth
Euderos
liyanqing
Mexico1604
Jenny199567
  • Oct 6
  • Switzerland
  • Deviant for 2 years
  • He / Him
Badges
Baby Blob: Achievement unlocked! Your 2nd deviation made the Mystery Egg hatch.
Bouncing Blob: Your 5th deviation's making it jump for joy!
Super Albino Llama: Llamas are awesome! (309)
Quartz: It's a big honor to be awarded a Quartz badge! (67)
Disco Blob: Groovy, baby! Your 10th deviation unlocked the funk.
My Bio

I love everything to do with the sky, clouds and especially airships and the steampunk fantasy

It's truly an amazing genera of imaginary dimensions

my work revolves around various powerful Ai models based on the legendary Katherine Crowson's cc module diffusion model for Ai generated art hence I cannot truly be considered an artist


This gig is just a page where you'll find everything related to fantasy scenery, quite often my generations are based upon fellow deviant artists, purchased images from stock websites and random images found on the internet.

Keep in mind that this gig aims in no way to monetize any recreated work or claim them as my own. It's just a gig to explore the capabilities of a computer to recreate by a base an artwork for educational purposes.

Hope you enjoyed my creations & recreations and take off on your own adventure 😊


I DO NOT TAKE COMMISIONS :cd:


Can I use your work : yes, however be warned that I take no responsibility for copyright violations as the images I re-creat are for single use & does not permit 3rd party reposting (even with consent)


Favourite Visual Artist
KateFox
Favourite Movies
Lord of the Rings, Harry Potter, Star Wars series
Favourite Bands / Musical Artists
Lofi Girl
Favourite Books
Lord of the Rings series
Favourite Games
Swbf2, Bf1, BDO
Favourite Gaming Platform
Pc
1 - Prompting As easy as it sounds, there's a lot more to a prompt than meets the eye as despite thinking it's just writing words as the challenge lies in articulating a concise and precise description which effectively "paints" the desired image while accounting for both the capabilities and limitations of the AI model. There's equally a very technical aspect to it too. I shall be using one of my personal favorite creations as an example The prompt for this specific image went along the lines of [((masterpiece))], ((best quality)) [ultra-detailed 4k highest quality][highres highly detailed], , solo focus , (close up shot) looking back, ((white hair)), (short hair),bob cut , ahoge, (squinting), constricted pupils, (half closed eyes), (grin, smile), pale skin, simple background, {{{lineart >> sketch}}} Keep in mind the above is the base prompt which created
Join the community to add your comment. Already a deviant? Log In
Since the 11th Nov, deviantart has become......a dumpster-fire or a Warzone Gulag with bullets and knives flying in all directions. As an ai model engineer who spent almost 6/7 months building his own model and being amongst the first 3 Ai generators on deviant art thought amongst the few who rework by hand the generations & receives death threats regularly by other actual artists I have to tell you by experience on how the models work that the "noai" flag will sadly almost not protect you at all Why? Think of it this way, I have an empty bucket (the model), I go to a shelf full of paint tubes (websites), take a handful of them (artworks) and empty them in my bucket (neural-network datasets),Each shelf has its different sets of tubes but a lot of them are the similar/identical so in the end I've got the same sets of paint tubes even thought I got them from different shelfs To put it simply, even thought you could opt your work out of a model, it's quite meaningless to the actual
Join the community to add your comment. Already a deviant? Log In
As many have noticed, DeviantArt is now host to a very rapidly growing community of various Ai Generating Artists and a few of them are not very well received. As I'm part of it, in this Journal I'll be explaining why and how I work. How does Ai Art work :la: Here I'll be covering the basic of the most common model Disco Diffusion : Disco Diffusion doesn't actually search the web for images to compose it's creations like other models such as Dall-E 2 for instance which does use an image decoder model, using an actual image for the generations to create a completely brand new image. Instead, neural networks that DD uses are pre-trained and use a noise generating model called Clip-Guided-Diffusion which creates an image through noise. The difference between Image and Noise :sherlock: The common argument for many against the concept of ai generations follows the theme of "stolen content" in the sense that an ai "steals" concepts and ideas from across the internet & artists to mash
Join the community to add your comment. Already a deviant? Log In

Profile Comments 56

Join the community to add your comment. Already a deviant? Log In

hey just wanted to say that all of your work so far has been amazing and even though I can't add them to my favorites due to the deviantart algorithm, they will always have a place there in my heart.

ai generated content is not welcomed here. please go somewhere else.

Up to the user to decide, if you have nothing to do and need something to hate to feel special then yes, but if your just here to look for wallpapers or images in general then no, besides all the stuff you see on my page is reworked/redrawn either in Ps, IL or both so there's equal effort into the images I upload then to your funny hand sketches, the model I made was trained on stock images I bought with a license so nothing was "stolen" either

Let me make something clear to you:


You are not an artist. You are using an image generator that is plagiarizing from the entirety of the Internet, to create a new image with absolutely ZERO effort on your part.