AI generated art

Elentarri said:
Not too shabby. I was expecting more 3 and 4 winged birds. :ROFLMAO:
Click to expand...
Just for you
strange bird.jpg



Meanwhile, getting more dynamic but it seems more down to luck than isolating the correct text to use. Still can't find a common denominator.
grob2.jpg
grob3.jpg
 
I don't think there is a common denominator - yet. I think the AI just spits out old images from the internet, semi-tailored to whatever text +style input you added. That's why the anime works best and the rest doesn't. Just a guess. And the AI is limited to the shitty google-search algorithms the rest of us have to deal with, as well. So it gets crappy graphics to start with. I'm sure the more images + style + input text there is, the more the AI improves (or someone else will write some sort of code to help it out). Different freebie/semi-freebie AI Art Generators may produce different graphics based on the same prompt. NightCafe has more options in terms of style and type of input than Muzzy. So, different end result.

There is, also, the possibility that the particular AI you got stuck with has a personality; likes anime and is willing to put in the extra effort. The rest is a half-arsed attempt. ;)
 
Stable Diffusion seems to only have enough umph for a certain amount of detail -- And the amount is less than the number of pixels in the images it creates -- It is super weird to me. Anime doesn't require nuances like skin texture, it SD can focus on other things. Here is an attempt at a Photograph style image.
1735887442015.jpeg

By the way, the term "Bokeh" is a photography term that AI seems to understand. I won't attempt to explain it here. It is worth looking up, and using if it is appropriate to the images you want.
 
I use a Lensbaby Composer, which I think is quite good for the bokeh effect.

lensbaby.com

Lensbaby | Creative Effects Camera Lenses for Photography and Video

Lensbaby creates award-winning creative effects camera lenses, optics, and accessories for photographers. Our Photography gear and equipment helps creatives create in-camera effects. With our selection of camera lenses for DSLR and Mirrorless cameras, photographers can find the tools they need...
lensbaby.com
 
JunkMonkey said:
Being bloody minded I sometimes just type in random strings into free online AI image tools; things like '76$#\+7)¥¢>{{{'.
It's surpring how often these chunks of garbage are interpreted as meaning 'show me an attractive underdressed Asian woman staring invitingly into the camera'.
Click to expand...
:ROFLMAO: :ROFLMAO: :ROFLMAO: :ROFLMAO: :ROFLMAO:
 
JunkMonkey said:
Being bloody minded I sometimes just type in random strings into free online AI image tools; things like '76$#\+7)¥¢>{{{'.
It's surpring how often these chunks of garbage are interpreted as meaning 'show me an attractive underdressed Asian woman staring invitingly into the camera'.
Click to expand...
this is very true on Muzzy using the various Anime styles.
 
Last edited:
I find as long as you use the PG-13 option, it keeps them to a minimum. The problem with that is that it also limits the options for 'Cosmic Horror' (one of my favourites to play around with).
 
JunkMonkey said:
Being bloody minded I sometimes just type in random strings into free online AI image tools; things like '76$#\+7)¥¢>{{{'.
It's surpring how often these chunks of garbage are interpreted as meaning 'show me an attractive underdressed Asian woman staring invitingly into the camera'.
Click to expand...
Nonsense is also a way to get more variety from a prompt you already like.

So if you like the prompt "Green Robot" you can try the prompts "Green Robot 001" and "Green Robot r4adf5a4r" and see if you get some variation.
 
It interests me for a time and then gets kind of boring. But then I see one that surprises me.

Those fake trailers--I saw this one for a Snow White movie.
The animation looks great but the composition and content action is just so repetitive and sterile.
Painful painful narration:

On the other hand
this one is pretty darn good in terms of feeling authentically retro:

 
AI is dumb. It has no idea that an image it has just created of a human with three hands or seven fingers is inaccurate.

It can draw on images from its databank when required, but has no concept of what they mean.

And long may that continue.
 
I tried to get an AI image generator to make a Blemmye but it came out with an anthropomorphic muskrat-looking thing.
I like some of the work here.

There's some gender-bending imagery and multiple fingers but some are quite nice:

 
paranoid marvin said:
AI is dumb. It has no idea that an image it has just created of a human with three hands or seven fingers is inaccurate.

It can draw on images from its databank when required, but has no concept of what they mean.

And long may that continue.
Click to expand...
Like any new technology, I think AI will be vastly overused for quite a while. It will be implimented in the most ridiculous ways. Perhaps an AI remote control that learns to anticipate your TV channel surfing habits. But, eventually, it will calm down and will be used where fast analysis or processing is required. An example could be that it becomes very useful in the automation of preserving and restoring old cinema classics.

It's already proving useful in medicine but I'm still unconvinced that it has any artistic merit. I have to admit that I do like playing around with it in the visual medium but, in mitigation, I do not consider myself, the AI or the results to be artistic in any way. It's just fun (which makes me guilty of the overuse I began this statement with):)
 
You must log in or register to reply here.
Thread starter Similar threads Forum Replies Date
Coragem New Amazon rules on "AI-Generated" Content. Publishing 9
S Why I think AI-generated art is not art Art 151
Mon0Zer0 AI generated video Technology 18
Mon0Zer0 Post a story based on this AI generated image (max length 500 words) - Sep 2022 Workshop 3
Valtharius AI Generated Stories Workshop 2

Similar threads


Back
Top