FA+
Shop
I tried it, I'm disappointed and unimpressed.
11 months ago
Hello again Fellow FA Fans, Friends, Artists, Writers, Poets, Fursuiters, Dreamers and other highly creative people.
These past few journals I was trying to relate my impressions concerning AI created art, and between the artwork created by "Conventional" means. I was really surprised and impressed by the responses I got and the results ended up pointing strongly to "Just keep doing what you do best." Well... My major disappointment was that even after following the instructions on the AI site closely, either I messed up on the inputting of instructions, or I couldn't get the algorithm right. I also had major concerns about giving out too much personal information as I had suspicions these AI art sites are prime Phishing Holes for cyber scammers. I also felt a bit enraged when after getting a near perfect image of my character "Muse" on the DA "Dream Up" page, the site demanded payment.
"Well Foo!" I decided. "I could draw it much faster! -Or at least I could get the ideas down in a more concise manner."
Actually, what I really need is supportive information and instructions from someone I could trust. One of my friends here on FA told me a story about how he almost slammed a very expensive and high-end race car into the haybales on his first try behind the wheel, and I couldn't help but feel a similar sensation after the DA experience. Perhaps this rendering method requires much more skill and understanding of how these systems work, as compared to my "Old Fashioned" ways of creating my hand drawn images and other artistic endeavors.
Let's face it friends. It's like I'm trying to keep up with the pack on the opening lap of a modern stock car race in a 1977 era Chevy Nova with a non-modified six-banger.
Anyway. I appreciate all the advice of all who responded, and I do hope there might come a time when I'll be able to render up the beautiful, semi-realistic backgrounds, all the gorgeous colors and the wonderful textures that the AI seems to create so easily. With Windows 11 looming over the cyber-horizon, and my current "Art Machine" still limping along on Windows 10 and updates, I've been contemplating on getting a fresh, new PC or perhaps an Apple machine to start over again. A LOT of my more computer savvy friends here on FA and my musician brother use Apple products. Perhaps it's time for me to take the next step up.
"Peace."
These past few journals I was trying to relate my impressions concerning AI created art, and between the artwork created by "Conventional" means. I was really surprised and impressed by the responses I got and the results ended up pointing strongly to "Just keep doing what you do best." Well... My major disappointment was that even after following the instructions on the AI site closely, either I messed up on the inputting of instructions, or I couldn't get the algorithm right. I also had major concerns about giving out too much personal information as I had suspicions these AI art sites are prime Phishing Holes for cyber scammers. I also felt a bit enraged when after getting a near perfect image of my character "Muse" on the DA "Dream Up" page, the site demanded payment.
"Well Foo!" I decided. "I could draw it much faster! -Or at least I could get the ideas down in a more concise manner."
Actually, what I really need is supportive information and instructions from someone I could trust. One of my friends here on FA told me a story about how he almost slammed a very expensive and high-end race car into the haybales on his first try behind the wheel, and I couldn't help but feel a similar sensation after the DA experience. Perhaps this rendering method requires much more skill and understanding of how these systems work, as compared to my "Old Fashioned" ways of creating my hand drawn images and other artistic endeavors.
Let's face it friends. It's like I'm trying to keep up with the pack on the opening lap of a modern stock car race in a 1977 era Chevy Nova with a non-modified six-banger.
Anyway. I appreciate all the advice of all who responded, and I do hope there might come a time when I'll be able to render up the beautiful, semi-realistic backgrounds, all the gorgeous colors and the wonderful textures that the AI seems to create so easily. With Windows 11 looming over the cyber-horizon, and my current "Art Machine" still limping along on Windows 10 and updates, I've been contemplating on getting a fresh, new PC or perhaps an Apple machine to start over again. A LOT of my more computer savvy friends here on FA and my musician brother use Apple products. Perhaps it's time for me to take the next step up.
"Peace."
That said, you certainly don't HAVE to play around if you don't want to, and as much fun as I've had it's certainly not as rewarding or worthwhile as real art. I see AI art as to real art as Guitar Hero is to actual playing. Is it real? No. Is it as rewarding? No. But sometimes you just wanna mess around for a bit and have fun, and not everyone has the time or inclination to buckle down and actually learn to play the guitar. There's nothing wrong with slapping some plastic buttons and rocking out, as long as you don't get delusions of grandeur.
And while there are those who can work the prompt incantations of AI to come up with things, I recon there's still a place for the artistic equivalent of "The Repair Shop" in both traditional and digital media.
Regarding the impending "Art Machine" upgrade, if you didn't want to take a third option to nuke Windows 10 on your current machine and install a Linux OS like Linux Mint (which I've been happily using for getting on to a decade), going the Apple Mac route would be your best bet. Mrs Mayfurr swapped out her old Asus Windows laptop for a brand-new Macbook Pro this year and she's very happy with it - even if Apple does keep prompting you to "go all the Apple Way" and buy more of their stuff
I still see people getting some pretty amazing results on DA with DreamUp, but I have to assume they're paying for the extra horsepower, which I refuse to do since I still regard AI-generated art as just this side of outright plagiarism.
That said, I can get some useful results, but most are limited to things like backgrounds and references. Occasionally I'll get something fairly nice out of Bing/Create, provided you can word your way around their various "adult" filters. I post that stuff on my Discord server. If you'd like to check it out, just let me know. In the meantime, I'll leave you with a link to one image I'm rather happy with, though it took a lot of post-processing in Photoshop:
https://www.deviantart.com/stash/023xz7xgn4bx
AI will always have a cold inhuman feeling, only content-consumers like it but they have no sense for art... believe me, people who are TRULY into art, don't find much charm in AI, aside from the technology itself! ;)
I can get that by opening my eyes and looking around. I want something new. Something that feeds my brain interesting things. Abstracts, cartoons,, distilled and creative reality.
But kudos to trying it, at least, and getting a feel for it yourself.
The problem is, if it's not trained on something, it can't really do it. An example I've used before is boobhats--if it's not trained on images of boobs resting on heads, in simply can't interpolate the individual parts into place. It can generate very pretty-looking details and textures (and yes, boobs), so there's a sort of instant gratification from seeing your words turned into graphics..but then you notice the hands like trees or piano keyboards, arms melded into tails, all the weird little things which should be so simple to fix with just the right words. Because your words can clearly do so much, it should be simple, right? And there's where the gambling addiction parallel kicks in.
I play with it for fun now and then, more with the attitude of it being like a video game than anything serious. I suppose if you wanted to run your own instance and train it completely on your own stuff, piling in scan after scan of your existing art, it might start coming out with stuff that's along the lines of what you're after in your prompts, maybe needing a little digital editing to fix the odd bit of weirdness.
As for Apple, see if you can borrow one for a while, even just parking yourself in front of a friend's at their place for several hours. I like it better than mswin, but that's partly 'cause I'm a unixhead and like having access to its command line (and yes to other readers, I've tried CygWin). It's a very different environment, and if you're used to Win10 and find you don't take to Mac-isms, you're probably best sticking stubbornly to Win10. I keep a couple Dell refurbs kicking around for times when I have to fire up mswin and assiduously avoid Win11 (likewise held off ever upgrading from XP then Win7 in turn 'til it was unavoidable). Time spent learning a new system is time spent not doing what you intend to use your system for, and at least with MS, they seem to be getting worse with each version since Win7. But remember:
https://www.youtube.com/watch?v=CPR.....MeMI&t=86s
Skype has Copilot integrated ,which can generate images.
https://designer.microsoft.com is their primary image generator. You have a stack of "fast" image generations free every day, afterwards the service gets ... sloooooooowwwww, but you don't pay for it. Though, obviously they will advertise their service as they want not only your data, but also your money. But they're Microsoft, so, when you paid for a Windows PC, they got both already anyway.
=> https://www.furaffinity.net/view/58657790/
Microsofts Image generator is "Instructions only".
The other alternative is Google's Gemini.
It can generate images since a few days. It obviously harvests your data, too, but at least Google anonymizes the data before selling it. Again, when you have an Android smartphone, Google has that data anyway.
Gemini you find here:
https://gemini.google.com/app
An example?
"Please, Gemini, generate the image of an anthropomorphic zebragirl with long black-white striped hair flowing down over her left shoulder, her eyes closed, dressed in a light green silken toga, kneeling on a blue cushion in an arabian garden, playing on a zither.
Her pose shall express hope and love."
=> https://www.furaffinity.net/view/58657726/
Google Image generator is "talk with a person".
Overall I agree with Reve: You can paint, draw & sketch.
Use that skill!
I can only do right angles.
protip: be careful with prompts like "fluid lineart"... we had a good laugh or two.
don't ask.
aside that, sometimes prompts are just ignored, for whatever reason. or the weighting is out of line...
I rather talk to a real human about what's to be done, and rely on their creativity.