FA+
Shop
To anyone with a passing curiosity about this AI art stuf...
2 years ago
General
Yes. Yes. Yes. I know. These digital demons are here to steal all our jobs and our hobbies and our incomes while they thieve from artists like the ultimate tracing pirates that they are. AND I fully support the view that here on FA, there is to be no direct AI image dumps, as that is a total waste of FA's precious bandwidth. These arguments on both sides are both very strong and valid.
I'm not here to dispute that the soulless army of art terminators are here to destroy the very foundation of art as we know it. Instead, I view this as a very powerful tool that can evolve to become great allies with artists. Besides, my art is writing, and the AI bots can churn out pages of stories to give me heartburn if I let it. Nor have I got the excess money sources to commission art pieces with. In fact, my primary goal with these new age tools is to bring forth Plight into a visual novel... maybe even a TV grade series that looks like an indie CG movie on the budget of "Got change?" (Can rattles). Unlike Fred with his amazing love towards 'A Fox in Space', who is putting in decades of work for half an hour of anime, I'm pretty certain that I shall never lay down a single frame of animation. So, that is where I am coming from. It's this AI stuff, the world's biggest winning lottery ticket or... basically nothing.
Alright. Rant over. Now, as I was saying, for those with even a casual curiosity about the new AI progression scene, here is a downright amazing offering.
The tutorial on how to do this can be found at this YouTube video...
For now, the entry requirements are just slightly an exclusive club.
- Windows. (Modern)
- NVDIA Graphics card.
- Minimum 4 GB of dedicated video memory.
- Lots of HDD space.
It's all explained in the video.
Check mark those and congratulations, you can opt in to have a MidJourney grade art maker, totally for free and totally running locally on your own computer. No fees per art piece. No credit limits. No monitoring staff to sensor what is being made. No... you get the point.
One downside is that since this is being run on your hardware and not a server, the production speed is at the mercy of the hardware specs you have available. For me, I want to say it was roughly 15 minutes (probably more like 10) for two images on 6 Gigs of V-Ram. However, have some movie time while messing around with this would be a perfect blending.
I tried for a cat in a tree. Got a really nice black & white with some slight floating superpower... comfortably hanging onto a limb and a tabby with some kind of super weight gain issues.
Then I tried for some fox images. Running through a garden was really lovely. Reclining in a bowl just wowed me.
Furries at a convention. Random species suiters, standing, in faux fur inside some brown-ish walled conventions spaces. A blind person could tell this isn't a photo. But dang, the uncanny valley is there. It knew how to do that scene.
Thought I'd trick it with 'Pac-Man and ghost in a maze'. Ehhh... round, yellow ball with eyes, a slight mouth. The ghosts were iconic. And the rooms were maze decor. I thought it would totally do something far more random.
I tried for a spaceship in orbit. And my sci-fi side half expected Star Trek ships. Oh, wow. Got a two-tier sphere with tons of details, landing on a sandy planet with a suited explorer already on patrol. That scene could rival any pulp 50's image with details, galore.
I also had a little time to try a couple other things. But what this thing produced was way above my expectations.
Now... most every time that i have commissioned an art piece for one of my literary visions, I have been extremely satisfied with the end results. It's like this magical connection of my thoughts have truly been relayed and created by the Human hand that is far more talented than my stick figures could ever be.
These AI images... they wow me with what they are capable of doing. Yet, trying to tell one what I see is like an experiment in trying to create a headache. In a few years, that may not be the case. However, for now, it is a far better novelty to glimpse into what is to come then to argue that these things are going to take over every aspect of life. Sure. They are autistic tracing Gods. But then again, I never dreamt of a starship captain's quarters with lush bed, workspace and such like this one AI image produced. (It's on my virtual wallpaper, right now.) Or the corridor that leads to it with a sky window strip for viewing the stars through. Now that is pretty darn creative... for virtual dice.
Over these last few days, I've been working with this image creation tool. Got to say that what it is doing is pretty impressive. Right now, I've found a way to make it pump out standing foxes in white dresses... which is a starting point towards my novel named Plight. Maybe with that and some DeepFake voice software, I can even get some kind of sample work started for a chapter or two that goes beyond just text. And in a few years... actual video???
Thinking about compiling some of these foxy images into a YT video and linking. May be a loophole to the FA ban over everything AI made.
And now that I've been working with this for over a week, I am even more impressed. Sure. I got a fox passing through a bathtub. And a lion missing a leg. But for AI oddities, those are the exceptions. Many weeks ago, those were the common problems. This one still can't spell. But that's alright. It's been staying fairly consistent for continuity and pumping out quality images... even if they still look computer made. The faux creativity is insane. It is listening... mostly... to my key words and doing its best to make the images work. But I can tell that this thing is heavily favored towards the G rated stuff. They'll hold paws but seem to be otherwise avoiding bodily contact for the most part. They'll sit doggy style. But it is all so smooth down there. I can get tails to raise. But the thing is too shy to do the work from behind when that is happening. I've gotten some prelude to kiss images. They'll close their eyes and get really close. It's like just a second away from happening. It does know how to open their maws. So, I tried to put a mousie in there. The thing totally drops the notion. However, like what I got it working on for gardening scenes, it's quite happy to give me standing foxes in white gowns that do the work. I put in 'Human' as a negative prompt. Yet a man and a lady are filling a basket with... something green. So, it is being hit or miss. And doing this on a home computer is slow. Minutes per image over seconds from a server. At least it's not paying for the privilege of that hit or miss approach. I also got it to produce a few 'gaming room' shots. To this AI, board game means chess. Although I was hoping to sneak around the no contact stuff with a game of Twister. Got an interesting table set that reminds me of Sit 'n' Spin with a bunch of colorful dots and two Foxes kneeling while playing the game of chess. The thing really knows how to make things coherent on where parts should be going together seamlessly.
I may have enough sample pictures to go ahead with a practice micro story about Camp Foxy, which will be a place for rich ladies to go for a life changing experience. And maybe I can even get a sample beta presentation of Plight going with this thing. Got some gardening shots that show promise. Now I need some somewhat consistent village sets to be produced. It will go over to one of my YT channels. Preferably one that has zero followers on it.
HOWEVER... I still would say that if Human artist styles are like comparing apples to oranges, AI Art is currently a kumquat. It is its own thing and right now, it is doing its own thing. In a few years, this thing will probably be reading our thoughts... when given the chance. For now, it's taking some educated guesses at what reality even means.
I'm not here to dispute that the soulless army of art terminators are here to destroy the very foundation of art as we know it. Instead, I view this as a very powerful tool that can evolve to become great allies with artists. Besides, my art is writing, and the AI bots can churn out pages of stories to give me heartburn if I let it. Nor have I got the excess money sources to commission art pieces with. In fact, my primary goal with these new age tools is to bring forth Plight into a visual novel... maybe even a TV grade series that looks like an indie CG movie on the budget of "Got change?" (Can rattles). Unlike Fred with his amazing love towards 'A Fox in Space', who is putting in decades of work for half an hour of anime, I'm pretty certain that I shall never lay down a single frame of animation. So, that is where I am coming from. It's this AI stuff, the world's biggest winning lottery ticket or... basically nothing.
Alright. Rant over. Now, as I was saying, for those with even a casual curiosity about the new AI progression scene, here is a downright amazing offering.
The tutorial on how to do this can be found at this YouTube video...
For now, the entry requirements are just slightly an exclusive club.
- Windows. (Modern)
- NVDIA Graphics card.
- Minimum 4 GB of dedicated video memory.
- Lots of HDD space.
It's all explained in the video.
Check mark those and congratulations, you can opt in to have a MidJourney grade art maker, totally for free and totally running locally on your own computer. No fees per art piece. No credit limits. No monitoring staff to sensor what is being made. No... you get the point.
One downside is that since this is being run on your hardware and not a server, the production speed is at the mercy of the hardware specs you have available. For me, I want to say it was roughly 15 minutes (probably more like 10) for two images on 6 Gigs of V-Ram. However, have some movie time while messing around with this would be a perfect blending.
I tried for a cat in a tree. Got a really nice black & white with some slight floating superpower... comfortably hanging onto a limb and a tabby with some kind of super weight gain issues.
Then I tried for some fox images. Running through a garden was really lovely. Reclining in a bowl just wowed me.
Furries at a convention. Random species suiters, standing, in faux fur inside some brown-ish walled conventions spaces. A blind person could tell this isn't a photo. But dang, the uncanny valley is there. It knew how to do that scene.
Thought I'd trick it with 'Pac-Man and ghost in a maze'. Ehhh... round, yellow ball with eyes, a slight mouth. The ghosts were iconic. And the rooms were maze decor. I thought it would totally do something far more random.
I tried for a spaceship in orbit. And my sci-fi side half expected Star Trek ships. Oh, wow. Got a two-tier sphere with tons of details, landing on a sandy planet with a suited explorer already on patrol. That scene could rival any pulp 50's image with details, galore.
I also had a little time to try a couple other things. But what this thing produced was way above my expectations.
Now... most every time that i have commissioned an art piece for one of my literary visions, I have been extremely satisfied with the end results. It's like this magical connection of my thoughts have truly been relayed and created by the Human hand that is far more talented than my stick figures could ever be.
These AI images... they wow me with what they are capable of doing. Yet, trying to tell one what I see is like an experiment in trying to create a headache. In a few years, that may not be the case. However, for now, it is a far better novelty to glimpse into what is to come then to argue that these things are going to take over every aspect of life. Sure. They are autistic tracing Gods. But then again, I never dreamt of a starship captain's quarters with lush bed, workspace and such like this one AI image produced. (It's on my virtual wallpaper, right now.) Or the corridor that leads to it with a sky window strip for viewing the stars through. Now that is pretty darn creative... for virtual dice.
---===---Over these last few days, I've been working with this image creation tool. Got to say that what it is doing is pretty impressive. Right now, I've found a way to make it pump out standing foxes in white dresses... which is a starting point towards my novel named Plight. Maybe with that and some DeepFake voice software, I can even get some kind of sample work started for a chapter or two that goes beyond just text. And in a few years... actual video???
Thinking about compiling some of these foxy images into a YT video and linking. May be a loophole to the FA ban over everything AI made.
And now that I've been working with this for over a week, I am even more impressed. Sure. I got a fox passing through a bathtub. And a lion missing a leg. But for AI oddities, those are the exceptions. Many weeks ago, those were the common problems. This one still can't spell. But that's alright. It's been staying fairly consistent for continuity and pumping out quality images... even if they still look computer made. The faux creativity is insane. It is listening... mostly... to my key words and doing its best to make the images work. But I can tell that this thing is heavily favored towards the G rated stuff. They'll hold paws but seem to be otherwise avoiding bodily contact for the most part. They'll sit doggy style. But it is all so smooth down there. I can get tails to raise. But the thing is too shy to do the work from behind when that is happening. I've gotten some prelude to kiss images. They'll close their eyes and get really close. It's like just a second away from happening. It does know how to open their maws. So, I tried to put a mousie in there. The thing totally drops the notion. However, like what I got it working on for gardening scenes, it's quite happy to give me standing foxes in white gowns that do the work. I put in 'Human' as a negative prompt. Yet a man and a lady are filling a basket with... something green. So, it is being hit or miss. And doing this on a home computer is slow. Minutes per image over seconds from a server. At least it's not paying for the privilege of that hit or miss approach. I also got it to produce a few 'gaming room' shots. To this AI, board game means chess. Although I was hoping to sneak around the no contact stuff with a game of Twister. Got an interesting table set that reminds me of Sit 'n' Spin with a bunch of colorful dots and two Foxes kneeling while playing the game of chess. The thing really knows how to make things coherent on where parts should be going together seamlessly.
I may have enough sample pictures to go ahead with a practice micro story about Camp Foxy, which will be a place for rich ladies to go for a life changing experience. And maybe I can even get a sample beta presentation of Plight going with this thing. Got some gardening shots that show promise. Now I need some somewhat consistent village sets to be produced. It will go over to one of my YT channels. Preferably one that has zero followers on it.
HOWEVER... I still would say that if Human artist styles are like comparing apples to oranges, AI Art is currently a kumquat. It is its own thing and right now, it is doing its own thing. In a few years, this thing will probably be reading our thoughts... when given the chance. For now, it's taking some educated guesses at what reality even means.
