This software utilizes an artificial intelligence dialogue system called ‘Buddhabot’ – Copyright AFP Behrouz MEHRI
The old “Is it art?” debate is going nowhere. It has been around for at least a century. As rhetorical questions go, it’s one of the least helpful, except for sarcasm.
AI art has been criticized by many for being “cheating, uncreative, and derivative”. Frankly, so what? Formal graphic arts is nothing new, especially in commercial sales.
Then there is another visual cultural atrocity, the so-called consumer art. These are real splashes of color, paint on her wheels, etc. What these ordinary pieces have in common is that they are mass-produced.
All you need to do is repeat the visual formula and sell it. Rather annoyingly, it can be the only way to make a living. Unless you’re an incredibly wealthy person, a money launderer, or a syndicate buyer, people typically don’t buy original, creative artwork.
…so why the growing outrage against AI art? Check out some random AI art samples we found online. We omit the technical details. If you know even a little bit about RGB or CMYK palettes, you know everything there is to know. What you may not know is the range of color definitions, tones, and palette options.
The sample linked above is a good example of the current limitations of AI art. I’m also aware of the style, but the palette is pretty simplistic – AI art doesn’t seem to stray away from the basics of color.
3D styles have also not “evolved” beyond the baseline art student level. AI is a pretty good student, but it has problems with depth and definition. Color differences are a difficult subject, but so are image resolutions.
For example, some images often have the same foreground and background color. Line definition is mechanical. Gradients can be improved and made more versatile. That’s important for lighting.
The end result looks like something a caffeinated Adobe Illustrator could do in their spare time.
This is just another form of mass production. I have another problem. The “generator” process uses the language base to define the subject. There are dozens of different AI image generators available. Each is associated with a language base. So the literacy of the image seeker and the literacy of his AI generator becomes a big issue.
Nuance is difficult to achieve in this environment. What would it be like to “imitate butterflies in a bright spring field”? Just because you can generate an image this way doesn’t mean you’ll like it.
… elegantly introduces another topic for this article. What AI Art Can Do Giving a computer a shopping list of images, music, or literature is one thing.By definition what it is won’t do create outside the framework.
Synthesize. interpret. That’s one of the problems. To get the image you want you probably need to include the specific image you want to use. You need to show the AI which image you want to include in your image.
You are already back in the creative arts. You need to create AI-friendly images. At this stage of AI art, we can mass produce, but not certain elements of design.
AI art should be viewed as a useful derivative of artistic tools. Being able to do that is incredibly valuable. There is no good reason to complain about not being able to do that.
Yes, it’s art in the same sense as a paintbrush.isn’t it cause That’s what makes the difference.
______________________________________________________________________________________
Disclaimer
The opinions expressed in this article are those of the author. They do not reflect the opinions or views of the Digital Journal or its members.
