AI imagery; is it, or can it be considered art?

1705863792234.png1705863833387.png
I just asked copilot "make some album cover art that looks like a 70s sci fi book cover. the artist is continuum and the album is called fadeout" and this came out in 3 seconds
 
Continuum said:
I just asked copilot "make some album cover art that looks like a 70s sci fi book cover. the artist is continuum and the album is called fadeout" and this came out in 3 seconds
Click to expand...
Bonkers. It's even got worn sleeves.
I guess a relevant question is, could you afford to pay a real artist to do that kind of work for you instead?
 
i know of a few 'psy' visual artists who are pretty upset by AI. Its taken their idea* , simplified it, made it available to the masses and subsequently done their income in

*using a computer to generate imagery.

Im still to read a piece of AI literature that comes close to my favourite writers.

Dunno about music. I seem to think that the line where pooters/hardware stop and artists begin became blurred in the 90s when all the electronica/breakcore people went deep into that rabbithole and produced tonnes of amazing music. I dont give a shit how electronic music# is produced; if i like it, i like it

#= electronic music meaning dance/electronica stuff. Not folk or rock etc

Overall though, a machine cannot infuse its art with meaning. So if you are looking for art that says something beyond the superficial then you would need a human in there somewhere. Still, there is nothing stopping an artist adding meaning to AI art after the fact, but thats not quite the same as writing from the heart, imo

But then again, most people dont tend want meaning in art. They want either the latest trend, or for it to look good/pretty/whatever
 
Last edited:
Art is applied imagination - visually, sculpturally, theatrically, literary, musically etc. In other words, a conscious, deliberate and most importantly human process rooted in human experience. One may engage certain tools in order to achieve a particular outcome, but the tool itself is not the creator. If someone were to accidentally drop a violin causing the strings to sound, this is not music because nobody willed it.

Of course, one may create art by involving elements of chance - in music academia, this is known as indeterminacy or aleatory. A good example is Earle Brown's graphic score December 1952, which pushed the boundaries of composition in not assigning precise musical features such as pitch or rhythm like a traditional stave. However, he still makes creative decisions when constructing the score, such as allowing performers to determine these components, who in turn make creative decisions when interpreting it. Hypothetically, if someone deliberately dropped the violin as part of an experimental piece, they too are making an artistic choice and it now becomes something meaningful, though whether one would describe the results as music rather sound or performance art will depend on one's definitions of all three. The crucial constituent is intentionality.

I apply the same rule to AI. It can be a tool like any other, but never the artist, because computers cannot think and feel - they respond to programming. Ergo sounds/pictures/words generated independently of human expression, whether at the composition stage or in the manner of their arrangement, are not art by definition (mine), no matter how beautiful or thought-provoking they might be on their own terms. I might notice an ancient tree, struck by its form and colour, but it does not become art until I take out my camera and subjectify it.
 
Of course a machine can not infuse it's art with meaning, but a machine can't make art without human input in the first place.
I don't know much about the prompting languages that are being used to generate these images, but I would imagine that they are becoming increasingly sophisticated, as the AIs do, and so I see no reason why the human who becomes fluent and develops good communication skills in the language, can not infuse the art they create using the machine, with the meaning that they intend.
I think the pictures in my OP show this.
 
Im not an art critic, but i would say that the meaning is put in by the artist by the process of the brain/consciousness/self-awareness using physical tools, acting on a physical medium. i would say that software taking bits of this that and the third and making an image isnt consciousness using physical tools acting on a physical medium, but i might be wrong. The Starry Night for example is carved out on the canvas and reflects how the artist was feeling at the time. Obviously it goes deeper than this, but as i say im not an art critic. Nor do i care much for convincing anyone that im right.

I would say AI imagery is art though. I would go onto qualify that its simulated art, but art none the less. I would question whether simulated art could go any deeper than 'oh that looks good', as i think the simulated part means that it can only ever be superficially appealing. WHich, i might add, is more than enough for most people, and more than enough for it to be art.

I dont think prompt commands can ever create the sort of meaning in the way im obviously failing to explain. Partly cos my heart isnt really in explaining it. The reason being is i could spend hours knocking out a long arsed reply, and someone could quite simply accuse me of talking bollocks. Its all about knowing your audience.
 
Last edited:
Its a Computer program. The artistry lies with the programmer. However many trillion images it can generate an hour there’s bound to be some decent ones.
 
As someone who has had to become familiar with how to use Large Language Models from a work perspective, I can confirm there is skill required in developing dialog with such models that results in an intended outcome. So in that sense, there is intentionality if the human side of this partnership has a vision of what they intend to create at the outset, even if that vision is modified as a result of the iterative output of the model. It is a very iterative process working with these models, and prompts will often become quite abstract and unintelligible by the end. I kind of think of it like colour mixing where as the human you’re navigating this liminal space of concepts rather than colours - adding a dab of this and a smidge of that until you get to a visual representation of what you’re trying to express. There are of course less invested ways to use the tool, just like someone who needs to paint the sea, squeezes out a bit of ‘blue’ and whacks it down on the canvas and calls it a day.
 
You must log in or register to reply here.
Back
Top