AI-generated literature provides an “other” perspective, breaking and supplementing the limitations of the human point of view. Photo: TUCHONG

In recent years, advancements in deep learning technology and the rise of LLM-based chatbots like ChatGPT have sparked growing and often contentious debates about AI’s role in literary creation. On one side, discussions center on the profound impact of these technologies on literature, while on the other, scrutiny falls on AI’s creative mechanisms and outputs, with comparisons drawn to the strengths and limitations of human literary endeavors. 

If we move beyond fixed mindsets and transcend the constraints of anthropocentrism, viewing AI-generated literature not as an extension of classical traditions but as an experimental form of avant-garde art, we may, by examining it through the lens of a post-human era, uncover a richer and deeper understanding of its unique characteristics, controversies, potential, and challenges. 

Debate on AI literature 

A common attitude toward AI-generated literature today is to treat AI as a competitor, evaluating its works against human literary standards and norms to determine whether they qualify as “real” literature. Under such scrutiny, AI’s literary experiments have sparked both considerable debate and criticism. On one hand, people are astonished by AI’s increasingly sophisticated ability to mimic literary styles. On the other, its perceived shortcomings in comparison to human works are frequently highlighted. 

In particular, the rise of neural network-based AI systems like ChatGPT has significantly enhanced AI’s fluency, coherence, and stylistic consistency. However, its perceived deficiencies in areas such as emotion, inspiration, and creativity have become the focal points of critique. Such criticism, however, reflects a literary conception that elevates emotion, inspiration, and creativity as the hallmarks of literature—concepts rooted in specific historical periods and technological conditions rather than universal or eternal literary truths. 

AI-generated literature, as an emerging artistic form driven by technological advancements, inevitably challenges traditional literary norms, prompting a reevaluation of literary perspectives in response to evolving times and contexts. By transcending essentialist views and recognizing AI literature not as a flawed imitation of an ideal in crisis but as an experimental form shaped by new technological realities, we can better understand its unique characteristics and potential. Just as photography once reshaped the way we think about painting, AI literature invites us to reexamine and refine traditional literary concepts. 

One of the most common criticisms of AI-generated literature is its perceived lack of emotion. Rooted in Romanticism, “emotion” has long been a cornerstone of modern literary concepts, underpinning the expressive theory of literature—the notion that authors channel their emotions into their works to create resonance with readers. In response, many AI researchers have sought to simulate emotional expression. Yet, even beyond these technical endeavors, the debate over AI literature’s inability to convey genuine emotion remains highly contentious. 

In the course of its development, modern literature saw a gradual departure from the expressive theory of emotion from the mid-19th century onward, a shift that grew more pronounced in the various literary movements of the 20th century. This evolution led mainstream literary thought to shift steadily from an expressive to an objective theory. 

It is also worth noting that when engaging with literary works—whether created by humans or AI—the reader interacts directly with the text itself. The expressive theory presupposes that meaning from the author’s external act of expression, suggesting that emotions are first experienced by the author and then embodied in the text. 

Another common criticism of AI-generated literature is its perceived lack of inspiration and creativity. Romanticism placed great emphasis on “inspiration,” “creativity,” and “imagination,” forming the core of its “genius theory.” This reflects a deeply ingrained modern cultural belief that inspiration is the true source and value of art. However, when this perspective is applied to AI literature, complications arise. 

Discussions around emotion, genius, inspiration, and creativity ultimately lead back to the question of authorship. Modern literary concepts, grounded in Romanticism, are fundamentally rooted in an author-centric view. Criticism often focuses on whether AI, as the creator of a text, can provide the same utility and function as a human author. However, when we examine the history of literary concepts, we see that the relationship between the author and the text is not static. It is significantly influenced by the evolution of human technologies and media. 

At times, the author’s central role in literature has been closely tied to the widespread use of writing and printing technologies. However, with the advent of the digital age and the profound impact of digital culture on print culture, the role of the author is inevitably changing. This shift is drawing the human role in creation closer to, and increasingly aligning it with, the role of AI in literary production. 

AI potential in post-human perspective 

In the subsequent theoretical developments of American scholar Donna Haraway and others, the cyborg emerged as a representative figure of modern human existence. As humans enter an era of increasingly advanced networked information, the cyborg has taken on greater informationist connotations. The cyborg state of being involves not just a fusion of humans and machines but also a higher unity formed through the flow of information between them. At the same time, this concept challenges the traditional human-centered view, disrupting the idea of an independent, autonomous individual. 

In response to this post-human condition, we must avoid two extremes: first, reducing machines to mere passive tools, thereby neglecting their transformative influence on human existence; and second, granting machines absolute dominance, assuming they will replace humans as the new subjects. Instead, a more nuanced approach would involve reassessing the significance of human-machine integration, beginning with the non-hierarchical fusion embodied by the cyborg and exploring the opportunities and challenges it presents for human existence. 

In fact, within the technological framework of AI, there are distinct paths rooted in both anthropocentric and cyborg discourses. Drawing on J.C.R. Licklider’s distinction between Turing-style AI, machine-enhanced humans, and human-machine symbiosis, attitudes toward AI-generated literature can be categorized into three approaches. 

One perspective sees AI literary creation as akin to human authorship, envisioning AI as possessing the same core abilities—emotion, inspiration, and genius—as human creators. However, this view clashes with the principles of deep neural network machine learning, which is based on statistical analysis of human language texts, making probabilistic predictions about context rather than exhibiting genuine creative or emotional insight. 

Another approach treats AI as a supportive tool for human literary creation. In this model, both AI and human contribute to the final work, but the text is ultimately shaped by the human author’s intentions, with AI serving only as a means to enhance the realization of those ideas. 

A third view positions AI-generated literature as the result of a collaboration between humans and AI. Here, AI takes on a more active role than a mere tool in the creative process, but it does not fully assume the role of the author. Whether through the sequential or connective paradigm of AI literature, though the AI autonomously generates the text, human involvement remains essential throughout the entire creative process. Often, people focus on the AI’s role as the author in literary works, overlooking the crucial role humans play in the overall process of creation. 

When we consider the entire process of literary creation, acknowledging both the autonomy of AI and its reliance on human input, AI-generated literature can be more accurately viewed as a decentralized form of “cyborg literature.” It is not the self-expression of an AI subject, but rather a new form that aligns with a human-machine symbiotic information system, where natural language processing technology enables seamless communication between carbon-based humans and silicon-based machines. 

‘Cyborg literature’ 

Interpreting AI-generated literature as decentralized “cyborg literature” enhances its significance, offering a more nuanced understanding of its potential. As a form of non-centralized “cyborg literature,” AI literature embodies a broader expression of “human capability,” vastly expanding the scope and forms of literary expression. While such potentials were already visible in earlier digital and online literature, AI-generated literature has propelled them to new heights. 

When we consider AI-generated literature as decentralized “cyborg literature,” several key “crossings” emerge that highlight its significance. First, there is the crossing of human and machine perspectives. The significance of AI-generated literature lies in its ability to provide an “other” perspective, breaking and supplementing the limitations of the human point of view. Through this dialectical relationship between AI literature and human experience, we gain an opportunity to reflect on ourselves through the lens of the other. 

Second, there is the crossing of author and reader roles. In modern literature, shaped by print culture, authors and readers are typically seen as independent, isolated individuals, and creation and reading are considered personal acts. However, in the feedback loop of AI-generated literary activities, the identities of author and reader become increasingly blurred. Both humans and algorithms share roles in authorship and readership, meaning the cyborg-like integration of humans and algorithms forms the “author-reader” relationship. 

Third, there is the crossing of linguistic and non-linguistic texts. In the digital age, boundaries between artistic forms like literature, painting, music, and film are increasingly blurred, with a growing emphasis on multimedia and multimodal integration. As a product of new technology, AI-generated literature naturally exists in a digital form, inherently multimedia and multimodal. Presenting AI literature solely as linear written text is merely a constraint imposed by outdated conceptual frameworks. 

Li Guocheng is an associate research fellow from the School of Literature at Nanjing University. 

Edited by WANG YOURAN