AI Film Frenzy: Art or Imitation? Unraveling the Wes Anderson Mash-Up Frenzy
Curious Refuge recently gained attention for its AI-powered filmmaking. They created a series of movie trailers using a combination of ChatGPT, Midjourney, D-ID, and Eleven Labs. These trailers blended Wes Anderson's unique style with popular franchises such as Avatar, Star Wars, and Lord of the Rings. The trailers have been viewed almost 7 million times and sparked articles in The Hollywood Reporter, Billboard, and Business Insider.
Curious Refuge claims to be for "anyone who wants to utilize the latest [AI] tools to create incredible art." However, there has been some debate about whether these tools create art or even artists rather than mediocre content––a serviceable imitation of previous works–-by content creators. To answer this question, we must acknowledge that the trailers focus exclusively on Wes Anderson's aesthetics, which are distinct from other kinds and easier to unravel than subtle messages of grief, family, and loss (prevalent in Wes Anderson films). Consequently, the trailers look familiar but feel lifeless.[1] Complicating things is that these creators may not even be fans of Wes Anderson. The founder and CEO of Curious Refuge, Caleb Ward, may have added as much when he admits his idea came from ChatGPT, which likely borrowed it from much earlier Wes Anderson parodies like Patrick Willems's "What if Wes Anderson Directed X-Men?" which premiered in 2015 or an SNL parody from 2019 (which adds narration not present in the Wes Anderson films but featured in many subsequent AI-powered trailers).
Ward may also misconceive artists. In an interview with the School of Motion, Ward says that he "steals like an artist" and that we all take inspiration from others even if we can't remember the source of our inspiration. Such a remark suggests that human creativity and ingenuity are nonexistent. Instead, creativity only shows that you can't (or choose not to) remember where you get your inspiration. This is an artistic version of the blank-slate doctrine. Creative inspiration is often derived from various external sources, such as the natural environment, experiences, and technology. It can also arise from internal sources, such as memories, personal knowledge, and experience. However, internal sources like emotions, feelings, imagination, and fantasy cannot be recognized. Instead, they come from nowhere but our subconscious.
Nevertheless, Ward is correct in that most of us are inspired by others. For the artist this includes other artists and their works. That does not mean artists are stealing. Wes Anderson was inspired by the French New Wave movement, specifically by Hal Ashby, Francois Truffaut, and Jacques Demy. Anderson did not steal from these artists as much as he integrated their styles into his view of the world. Anderson's ability to weave these influences into something distinctly his own makes his work stand out even against his creative influences, and this is something that occurs in the journey toward self-expression. Artists bring their vision of the world to life through their creations. This isn't theft. Conversely, Ward demonstrates no self-expression. In fact, he subordinated the self to create undemanding content for easy consumption. Ward is a marketing executive, and his expression is growth marketing.
More broadly, Ward's attitude suggests we are just a statistical approximation of our experiences. Great artists steal or, in this case, steal like an artist is a common fallacy Clayton Pummill and I found throughout our research for Generative Artificial Intelligence: More Than You Asked For (available on Kindle, Amazon, Audible, and Apple Books). The idea is that generative technology is just a mathy version of what every creator does, which is to analyze the creative works of others to assist the creation of derivative works. The argument frames generative content as a statistical approximation of the training data, which it is in most instances. It extends further to the claim that generative technology is entitled to the creative works of others because humans steal, too.
However, such a comparison is a false equivalency, not an artistic axiom. The argument mistreats humans as statistical models who create approximations (statistical or otherwise) of their favorite artists. However, creators are not inferior versions of computers. The works of others do not influence them as a disinterested algorithm for a third party's benefit by providing creative content like training data to a computer. This is a poor attempt to use logic and reason to understand a world beyond logic and reasoning, including emotions, feelings, imagination, fantasy, and the subconscious. Besides, a statistical approximation by a machine learning algorithm is not consent or authorization to use copyrighted material, and existing laws punish creators who steal.
The philosopher Albert Borgmann distinguishes between commanding reality and disposable reality. Borgmann focuses on music, specifically the differences between stereos and instruments or "devices" versus "things." He writes, "A thing has an intelligible and accessible character and requires skilled and active human engagement. A thing requires practice, while a device invites consumption. Things constitute commanding reality, and devices procure disposable reality."[2] Sculpting, painting, or filmmaking is challenging to master, and the artist will invariably be limited in range because they have refined some skills, not others, and will be attracted to some projects, not all projects. However, generative technology (for better or [for] worse) is instantly available, undemanding, and invites thoughtless creation and consumption.
I have seen Pickleball by Wes Anderson, Dune by Wes Anderson, The Shining by Wes Anderson, A Nightmare on Elm Street by Wes Anderson, and Halloween by Wes Anderson. These examples ruin my enjoyment of Anderson's unique style and show that content is cheap to make, and the doubling down is a characteristic of growth marketing, not art. AI creates a disposable reality. Borgmann's distinction between commanding and disposable reality underscores a fundamental tension in our relationship with art and technology. As we grapple with the rise of AI and the ease of generative technology, we must consider our value on authentic human creativity and the meaningful engagement it requires, lest we succumb to a world dominated by disposable realities devoid of depth and personal connection.
That's what I think. What do you think?
[1] The best way of describing my feeling is a kind of jamais vu, which is obviously French and attempts to capture a strange sense of something we recognize but also happens to create a sense of unfamiliarity and novelty. Jamais vu is the opposite of déjà vu.
[2] Albert Borgmann, Power Failure, Grand Rapids, Mich. Brazos Press, 2003), p. 31.
