BreannaVinson; AI Art: Where does it leave the human artist?
AI art is an emerging scene that has only grown more prevalent
in recent years.
Such growth sparks debate over
the benefit, or lack thereof, of such
a development. Most notably,
where this shift concerns the human
artist. To further explore this
topic, the research article AI
Art and its Impact on Artists1 will serve as a guided exploration
of the impacts on artistic creatives, as well as a way to conceptualize AI art
as a whole.
The text opens by introducing the subject of conversation and identifying the impact of AI art as
harmful to artists. This clear perspective is held throughout and will be
discussed further in detail in the proceedings. Continuing, there is an identification of the system that creates
these AI images,
referred to as image
generators. These models’ creation is
described as having the intention of altering existing images in a way that
mimics another image. In short, with a collection of images to use as source
material, the AI aims to follow the patterns shown to make something that is
both an attempted recreation and a mixture of all those provided images.
As the output is not a one-to-one copy of
the source materials and is rather some combination of present aspects, one may question if
this can be boiled down to inspiration prompted by the source images. In short,
this is not the case. In fact, the very concept of such an ideal is shut down
rather quickly and is deemed to be a harmful perspective. The text argues that
this shifts “responsibility from the people and organizations that build these
systems”2 giving false agency to an artifact
that cannot achieve such. It is
clearly established that humanity is not to be given to a machine that is
inherently lacking in such, and the art created by it is to receive the same
treatment.
This point is furthered by defining what art is. Although notorious for having a variety of definitions, one is settled on for this discussion. Art is a collection of personal experiences, feelings, inspirations, and the interpretations of such into a creative form. The personal element of such allows for vastly different creations in the space. Katsuhiro Otomo’s Akira, and Claude Monet’s Nymphéas Bleus vary greatly in presentation, but both are clearly identifiable as art, upholding art’s core. This core is that of humanity, something AI art lacks. Ergo, it can only be representational of “true” art, lacking the soul of the work and leaving only the technical.
.jpg)
Cover of volume
no.1 in Katsuhiro Otomo’s series Akria
One of Claude Monet’s works from the collection Nymphéas Bleus
Not only is this an issue for those who care about art on a deeper level than mere aesthetics, but it quickly
becomes problematic for those who put their heart into their creative works.
Simply put, creating art from this
place of genuine care is no swift task. In addition to the years dedicated to
the craft, hours are spent to create
such works, possibly weeks or
longer. While this process is necessary for great art, it is unappealing to
employers. AI art does not require these long periods to create, nor is the
expense as high. This leads to it being seen as a shortcut to those who care
not for the depth of human- created works, looking only for a finished product.
Such a situation has clear implications regarding the state of artists, and it does not have a positive connotation. Job loss has already crept up due to AI usage, with image generators taking the place of human employees due to cheaper costs. “For instance, according to an article on Rest of World, a Chinese gaming industry recruiter has noticed a 70% drop in illustrator jobs.”3
This labor impact is of grave significance, as it stifles a creative’s ability to create professionally, an issue many already face. Only one of the multitude of issues, negative impacts such as oversaturation of illustrations due to AI image creation, continued representation of inaccurate and even harmful stereotypes, lack of discoverability for new artists, AI art does more damage than good to creatives.
| Four of Hollie’s illustrations used to train the AI model (left) and sample AI output (right) |
A prime example of this lies in the very training of these models. In order to create visually pleasing art, AI must already have art of this kind. These works must be taken from human artists, and there is currently a lack of regulation on how this is done. Multiple reports of artists discovering their work was used for such training without their consent are out there, and many find this not only invasive but a contribution to an inaccurate representation of them as creatives. Hollie Mengert remarks that she feels distant from results, not seeing herself in it, “it feels like it isn’t actually mimicking my style, and yet my name is still part of the tool.”4
So, what
does this mean for the human artist? Although the text provides ideas for
combatting these oppositions, they are
inconceivable for many and do not
address the root issue. With that,
I wonder if anything can be done to reach such a core. Those who support
AI models in favor of human artists share vastly different priorities than those who lobby
against such systems.
While I do not seek to make this too much of an
opinion piece, I cannot
disagree with those negative impacts
from the chosen text and see AI art ultimately as a threat
to creatives. There are also aspects that are still unclear, such as the impact
such a rise has on those in creative writing. The text briefly mentions this,
but does not delve into such, as there is a focus on visual arts.
1 Jiang, Harry H, Brown, Lauren, et al. “AI Art and its Impact on Artists.” ACM Digital Library. 29 August 2023. https://dl.acm.org/doi/10.1145/3600211.3604681. Accessed 6 September 2024.
3 Jiang, Harry H, Brown, Lauren, et al. pp. 367. 1
4 Baio, Andy. “Invasive Diffusion: How one unwilling illustrator found herself turned into an AI model.” WAXY.1
November 2022. https://waxy.org/2022/11/invasive-diffusion-how-one-unwilling-illustrator-found-herself-turned-
into-an-ai-model/. Accessed 10 September 2024.
AI in the art world is incredibly problematic, for many reasons, as both you and the article pointed out. It’s stealing jobs, actual artwork and creative ideas. As the article mentions, the art must be taken from somewhere, and that means stealing art from creators either willingly or unwillingly to feed the machine so it can use their art. This can be disheartening for artists at best, as it can look and feel detached from their original work, and harmful in almost any other case, as it steals work from unwilling artists. I agree that many of the ways to fight AI are hard to access and inconceivable for most people. I know this is a personal problem I have had, where I am not able to add filters like Nightshade or Glaze to my work as my computer cannot run such a process, and therefore I cannot protect my work from new AI policies on certain platforms.
ReplyDeleteI know you said that you weren’t seeking to make this an opinion piece, but I appreciate how you still worked to condemn AI slightly in your conclusion. I almost wish it was more of an opinion piece. I would like to hear more of your thoughts on this, and how it makes you feel, especially as a fellow creative. This is an important issue for us, and I would love to hear more from you. That being said, I fear how it's going to impact us as we leave school to go and join the workforce. AI is scary to me and feels like the antithesis of everything we are working towards here at this school, both on an individual level and as a whole student body. At the same time, I hold out hope that we’ll always value humanistic art over robotic kinds, and that the heart and soul we put into our work will draw people in and make our place in the world.
Jiang, Harry H, Brown, Lauren, et al. “AI Art and its Impact on Artists.” ACM Digital Library.
29 August 2023. https://dl.acm.org/doi/10.1145/3600211.3604681. Accessed 29
September 2024.