Is AI considered art? The Black Water Exhibition at the National Art Museum sparks controversy, with Guo Peiqi being criticized for trashy creations

Guo Peiqi, the artist behind the Heishui Exhibition at the National Art Museum, sparked a debate on Threads over AI-created works. Netizens criticized “AI Drawing Art World” as garbage, questioning the effortless gain. Guo responded that contemporary art is not limited to hand-drawing, and will showcase “Pink Battleship.” Netizens criticized structural collapse and finger errors. Some believe the debate itself constitutes art. MoMA has long collected AI works, asserting that the authority lies with art critics and museums.

The Structural Collapse of Pink Battleship and the Deadly Flaws of AI-Generated Images

Guo Peiqi immediately rebutted: “Contemporary art has never been only hand-drawn,” sparking further discussion. Facing criticism, Guo posted again, saying “too lazy to reply to every troll,” and claimed to be an invited AI artist for the “Heishui” biennale, showcasing the AI work “Pink Battleship.” The piece reimagines the father’s maritime voyage through AI, from a female perspective, deconstructing the traditional masculine impression of warships and warfare.

This further angered netizens, some questioning the quality of “Pink Battleship,” criticizing its AI-generated structure collapse, and pointing out typical AI errors such as incorrect fingers, distorted limbs, unreasonable perspective, and fuzzy object boundaries. These are common issues in current AI image models, even with advanced tools like Midjourney, DALL-E, or Stable Diffusion, which cannot fully avoid such flaws.

The criticism hits the mark: if AI works are to enter the art hall of fame, they should at least be technically impeccable. When human artists spend months or years refining a piece, their techniques, composition, and colors are repeatedly scrutinized. In contrast, AI-generated works, though faster to produce, cannot guarantee basic structural correctness. Why should viewers accept their artistic value?

Three Typical Flaws of AI-Generated Images

Anatomical Errors: Incorrect fingers, disproportionate limbs, joint distortions

Physical Violations: Unrealistic perspective, conflicting shadows, ignoring gravity

Detail Collapse: Blurred boundaries, repetitive textures, illogical local details

Some netizens pointed out that in the context of contemporary art, it doesn’t matter whether the artist personally made the work, citing Jeff Koons, Damien Hirst, etc., who design concepts and then delegate manufacturing to factories, not handcrafting themselves. This argument attempts to defend AI art but raises new questions: when Koons and Hirst commission factories, they can realize their designs perfectly; but with AI, the artist’s control over the final result is extremely limited, often with unexpected errors.

Another netizen claiming to be a graduate of the New Media Art Department at Beihang University called for keeping pace with AI’s rapid iteration and producing better visual effects to convince others that AI is art. They urged AI artists to approach this with humility, using innovative tools. This advice hits the core issue: AI is a tool, not an excuse. Using AI does not mean lowering quality standards.

Arrogant Remarks Spark Confrontation and Artist Identity Anxiety

Some argue that for young people aged 15-20 with “painting dreams,” the rapid generation of AI threatens their self-identity and skill barriers, potentially restructuring industry ecology and social labels. This anxiety is real. When someone spends ten years learning painting—mastering perspective, anatomy, color theory—and then AI produces similar or even more exquisite images in seconds, it’s understandable to fear skill devaluation.

Guo Peiqi previously claimed that “architects and interior designers will soon be unemployed” after AI model updates. Such arrogant statements intensify opposition. For architects and interior designers, this is not just a technical issue but a challenge to their professional value. When AI artists stand on the moral high ground of “technological progress” and declare the death of traditional professions, backlash is inevitable.

As the controversy continues, some discussions shift to a “meta” level. Some believe that viewing Guo Peiqi’s work alongside the online debate on Threads could itself be a provocative piece—“AI painting in pink disguise, like the black ships of the past…”. Others suggest exhibiting Guo Peiqi’s past arrogant remarks as well, arguing that this clash between AI artists’ arrogance and the public’s prejudice constitutes a highly discussable form of art.

This “controversy as art” argument is highly dialectical. If the National Art Museum exhibits both “Pink Battleship” and screenshots of this online debate, it would create a richer narrative layer. Viewers would see not only AI-generated images but also society’s real reactions to AI art. This reflexive curatorial approach could transform a quality dispute into a profound discussion on “art in the AI era.”

Who Has the Authority to Define Art? MoMA Has Already Answered

While netizens debate whether AI works can enter art exhibitions, many examples exist worldwide. Some argue that the true definers of art are critics, galleries, and museums—authorities. In fact, MoMA has long collected AI algorithmic works, and the New North Art Museum has exhibited AI-generated pieces.

This reveals the power structure within the art world. The definition of art has never been democratically decided but controlled by a few elites. When MoMA acquires a work, it automatically gains the label of “art,” regardless of public recognition. From this perspective, Guo Peiqi’s invitation to exhibit at the National Art Museum is an official recognition of his AI creations. Although critics’ voices are loud, they do not negate the institutional logic of art.

However, some netizens counter: “Although claiming that all AI-generated works are not art is an overstatement, from AI or even earlier—photography—when their emergence lowered the difficulty for the public to operate, does that also imply inviting the public to participate in the discussion of art?” This insightful rebuttal points out that when photography was born in the 19th century, it faced resistance from painters, who considered it “easy” and unworthy of art. Yet, after decades, photography was accepted as an art form, producing masters like Ansel Adams and Henri Cartier-Bresson.

Will AI art follow the same trajectory as photography? It depends on whether AI artists can develop a unique aesthetic language and methodology, like photographers. Simply pressing a generate button is not art, but if artists craft prompts, iterate, perform post-editing, and deeply understand AI models to create irreplicable works, they possess artistic uniqueness. Guo Peiqi’s problem may not be the use of AI itself but the quality and attitude of his work, which fails to convince viewers.