Wardlow 1
Lainey Wardlow
Professor Carlos Santana
PHIL 4010-001
6 May 2020
Artificial Intelligence and the Age of Optimistic Artistic Interdependence
In a recent interview for Sean Carroll’s Mindscape podcast, musician and producer Claire
Bouchier, known by her artist name Grimes, made the controversial assertion that Artificial
General Intelligence (AGI), or AI which possesses generalized sentient qualities, will surpass
humans in its capacity to “master the arts and sciences” within the next 20 to 30 years, a capacity
which she claimed would mean the “end of art, human art” (Carroll). Unsurprisingly, this claim
elicited strong emotional responses from indie musicians like Zola Jesus, who on Twitter
referred to Grimes as “the voice of silicon fascist privilege” (“Grimes”). Since the online
upheaval, Grimes has clarified her position, stating that she was just “riffing,” and that while she
remains optimistic about humanity’s artistic relationship with AI, it is nonetheless important to
consider the implications of a future in which AI is capable of creating art that is potentially
more beautiful or more engaging than human art.
Additionally, while Grimes may have brought this debate further into the mainstream
consciousness, concerns about AI surpassing humans in creative capacity are not particularly
new. Continuing the ancient philosophical debate around art and aesthetics, scientists and
philosophers are now using AI to explore whether creativity is an exclusively human trait and to
what degree that exclusivity undermines or enhances the importance of human art. On one side
of the argument, philosophers emphasize the capacity of AI to mimic the creative process, to
improvise and to create in their output works of art and music that are uniquely beautiful.
Wardlow 2
Although these researchers don’t suggest a connection between AI’s creativity and the
eradication of human art, their research certainly indicates that art may not be an exclusively
human specialty. On the other hand, some scientists and philosophers argue that, while AI may
be capable of a convincing art product, it is nevertheless incapable of meaningfully engaging in
the artistic process, a process which is just as important to aesthetes (i.e. those who appreciate
art), as the art itself, if not more so.
Overall the research demonstrates a couple of things: First, AI is mostly capable of
producing art and music that is indistinguishable from—and sometimes even preferable to—
human-made art and music. Second, physical beauty (i.e. the experience of art with the five
senses) is not the most significant aspect of aesthetic value. In addition to physical beauty, artists
and aesthetes alike value the artistic process, one which involves motive or intentionality as well
as what British mathematician Marcus du Sautoy refers to as “aesthetic cognition,” a process
which some argue is exclusive to humans or which, at the very least, must be entirely understood
by humans before it can be programmed into an AI (du Sautoy 75). That said, philosophers have
largely looked over what is perhaps the ultimate question with regard to AI and art: if AI
surpasses us in its capacity to create beauty and to participate meaningfully in the creative
process through both intentionality and aesthetic cognition, what does that mean for the future of
human art or art in general?
In this paper, I will delineate the three most frequently considered aspects of aesthetic
value—namely, (physical) beauty, intentionality, and aesthetic cognition—and argue that, while
it is theoretically possible for AI to match, if not surpass, humans in some of these areas, there is
no reason to believe that AI will eradicate the need for human art, even if these capacities were
to develop within Grimes’ proposed time frame of 20 to 30 years. Furthermore, I will argue that
Wardlow 3
although AI’s continued capacity to create certainly challenges our assumptions about
humanity’s inherent value or superiority to other life forms (including AI), this capacity actually
precludes significant advancements in human consciousness and creative abilities, so long as we
remain optimistic about our relationship to AI instead of attributing to it an intangible mysticism
that threatens our value.
First, given recent developments in AI’s ability to create art that satisfies our need for
physical beauty, many are concerned that this reduces, if not altogether eradicates, the desire for
human-made art. This anxiety isn’t particularly surprising, given that physical beauty is the most
obvious value of art: “obvious” not only in the sense that the product of the artistic process (i.e.
the art) is more visible than the artistry of the process itself, but it is also an obvious statement in
the sense that humans value beauty without exception. And although what constitutes beauty is
subjective, the desire to view, listen to, or engage with forms that elicit sensual euphoria is
arguably one of the only universal traits that humans share. The obviousness of this value in art
only makes the fact that AI is capable of eliciting these feelings in humans all the more
threatening. Furthermore, current research suggests that humans are either unable to reliably
distinguish computer-made art from human-made art, or they actually prefer the computer-
generated art to the human-made art.
However, the experience of physical beauty in art is also insufficient on its own. For
example, in a study done by Scheerlinck, Wagemans, Chamberlain and Mullins, participants
were asked to classify several art pieces into one of two categories—AI-made art or human-made
art—and then evaluate those pieces. The results showed that, while the participants showed no
reliable capacity to accurately distinguish art made by AI from art made by humans, they
nevertheless maintained a strong bias against the art they believed to be generated by AI. So, in
Wardlow 4
some cases where participants incorrectly categorized the art, they actually showed bias against
human art and toward computer-generated art. In other words, even in cases where participants
were looking at art made by humans, they maintained a bias against those art pieces if they also
believed them to be made by AI. This study shows that AI is capable of fooling its adversary in
an artistic Turing Test of sorts, which alone displays a respectable level of intelligence, but it
also shows that our bias toward physical beauty (i.e. visual, sonic, etc.) is largely rooted in our
connection to humanity.
On the other hand, this research also challenges previous notions that AI is artistically
impaired by an inability to incorporate spontaneity—a value integral to what we consider
beautiful and creative—into its artistic product. The previous belief was that beautiful art is not a
