ماشین خیال: تحلیل تأثیرات چت جی‌پی‌تی بر گفتمان ادبی و ساختارهای اجتماعی در عصر دیجیتال

Abstract
This interdisciplinary study investigates the profound impact of the ChatGPT language model on the production and reception of literary texts and its associated social transformations. The applied, descriptive-analytical research, combining theoretical frameworks from literary studies, digital sociology, and media studies, demonstrates that the emergence of AI text generation challenges our traditional understanding of concepts like authorship, creativity, and literary value. Employing a mixed-methods methodology—including thematic analysis, computational stylistics, Natural Language Processing (NLP), in-depth expert interviews, and a wide-scale survey—this research provides a systematic examination of this phenomenon. Findings indicate that while ChatGPT democratizes literary content production, it also leads to emerging phenomena like “algorithmic authorship” and “machine discourses”. Textual analysis reveals that while AI-generated works are formally acceptable, they lack the experiential knowledge and human insight of authentic literary works. From a social perspective, the study shows a deep generational divide: young readers (18-29 years) with an average acceptance rate of 4.5 out of 5 often cannot distinguish human-made from machine-made texts, whereas traditional institutions and older individuals (50+ years) with an average acceptance of 2.3 out of 5 show strong resistance. By proposing the “Distributed Agency” model, this study offers comprehensive policy solutions, including revising literature curricula to emphasize digital skills, formulating ethical and legal standards for AI use in literary creation, creating support mechanisms for writers’ livelihoods during the technological transition, and developing digital empowerment programs to reduce the generational gap.
Keywords: Distributed Agency, Algorithmic Authorship, Digital Literature, Computational Aesthetics, Machine Discourses, Artificial Intelligence (AI).
 
1. Introduction
Significant advancements in artificial intelligence, particularly in large language models (LLMs), have fundamentally reshaped the production and perception of literary content. The emergence of sophisticated generative models like ChatGPT raises profound questions regarding the nature of literary creativity, the evolving role of the author, and the future trajectory of the publishing industry. Developed on a Transformer architecture trained on extensive textual datasets, this technology demonstrates an ability to generate text of near-human quality (Brown et al., 2020), marking a transformative moment in literary history comparable in magnitude to the invention of the printing press (Eisenstein, 1980). This shift represents a transition of machines from passive production tools to active agents in the creative process, thereby challenging core concepts within literary theory—including “author,” “text,” and “reader”—and demanding their critical reexamination and redefinition in light of these technological capabilities.
This study addresses a significant gap in existing research by conducting a critical analysis of ChatGPT’s impact on contemporary Persian literature and its associated social structures. Focusing specifically on Iran’s unique linguistic and cultural context, the research moves beyond purely technical evaluations to explore the broader socio-cultural and aesthetic implications of AI integration into literary practices. By examining how global technological advancements interact with local cultural traditions, this investigation aims to provide a nuanced understanding that is both contextually grounded and theoretically informed. The study particularly investigates how AI tools are being adopted and adapted within Persian literary circles, how they are influencing creative processes, and how they are potentially reshaping the ecosystem of literary production, distribution, and reception in Iran.
Furthermore, the research explores the tension between technological innovation and cultural preservation, examining how Persian literary traditions—with their rich history and distinctive characteristics—are both challenged and potentially enriched by these new technologies. 
 
2. Materials and Methods
This applied, descriptive-analytical research was conducted using an advanced sequential-explanatory mixed-methods approach, integrating both qualitative and quantitative methodologies to provide a comprehensive analysis of the phenomenon. The study unfolded in three primary phases. The first, exploratory phase employed qualitative techniques, including thematic analysis following the Braun and Clarke (2006) method applied to a corpus of 200 text samples (100 human-authored and 100 AI-generated), alongside semi-structured in-depth interviews with 50 key participants from the literary field. The second, confirmatory phase utilized a quantitative survey distributed to 750 stakeholders, encompassing professional writers, literature students, critics, professors, and general readers, to gather broad, generalizable data. The final, integration phase involved the systematic synthesis of the qualitative and quantitative findings to develop a nuanced and holistic understanding.
The study’s statistical population included the 400 text samples and the 750 stakeholders. A combined purposive-random sampling strategy was employed, with the sample size for the survey arm determined using Cochran’s formula to ensure statistical robustness. Data collection instruments included a researcher-designed 45-item questionnaire, which demonstrated high internal consistency (Cronbach’s α = 0.89), a 15-item semi-structured interview guide, and a detailed text analysis protocol featuring 75 quantitative and qualitative indicators. The validity and reliability of these tools were rigorously established through expert review (Content Validity Index > 0.87) and statistical measures, including strong inter-coder agreement (Kappa > 0.79).
Data analysis was conducted on multiple levels. For the qualitative data, thematic analysis involving open and axial coding was performed. The quantitative data were subjected to advanced statistical analyses, including independent samples t-tests to compare human and AI-generated texts, one-way ANOVA to examine differences between stakeholder groups, exploratory factor analysis, and Structural Equation Modeling (SEM) to uncover underlying structural relationships. Furthermore, computational text analysis was carried out using Natural Language Processing (NLP) techniques. To ensure the highest quality and credibility of the research, various strategies for establishing validity and reliability were implemented throughout the process.
 
3. Data
The analysis of 400 text samples revealed statistically significant differences (p<0.05). Human texts were superior in Thematic Depth (Human Avg: 4.2 vs. AI Avg: 2.7), Use of Metaphor (72% vs. 31%), Empathy Induction (4.3 vs. 2.8), and Originality (4.4 vs. 2.5). AI showed a slight advantage only in Narrative Cohesion (4.1 vs. 4.5). The survey confirmed a deep generational divide: 18-29 years: Avg. Acceptance: 4.5/5, Regular Use: 78%; 30-49 years: Avg. Acceptance: 3.4/5, Regular Use: 45%; 50+ years: Avg. Acceptance: 2.3/5, Regular Use: 15%. Qualitative analysis identified four main themes: (1) Shift from author-as-creator to author-as-curator/editor, (2) Generational duality (tool vs. threat), (3) Ethical and legal concerns, and (4) New creative opportunities for hybrid genres.
3. Discussion
The integration of quantitative and qualitative findings reveals that advanced AI like ChatGPT is instigating a paradigmatic shift within the Iranian literary field. Through the lens of Bourdieu’s theory of cultural capital, this transformation signifies a redefinition of the resources conferring legitimacy: traditional cultural capital—rooted in mastery of literary heritage and creation ex nihilo—now contends with an emerging technological capital, comprising proficiency in wielding digital tools, guiding algorithms, and curating machine-generated content. This shift transcends technical evolution, sparking an identity revolution that forces writers, critics, publishers, and educators to renegotiate their positions and values within the reconfigured literary landscape.A stark generational divide underscores this upheaval, with acceptance rates of 4.5 among youth versus 2.3 among older adults. Qualitative data reveal this is not merely a digital gap but a clash of aesthetic habitus. Younger writers exhibit a pragmatic, postmodern sensibility, viewing literature as content to be optimized and AI as a legitimate tool for brainstorming and productivity. In contrast, older generations adhere to a romantic-modernist worldview, venerating the solitary artistic genius, original self-expression, and the sanctity of creative struggle. They perceive AI as sacrilege—a threat to the unique authorial voice.
This tension necessitates re-examining core literary tenets. The classical “solitary author” model is evolving into “distributed agency,” where creativity emerges from human-machine collaboration. The author’s role shifts from sole creator to creative director—curating, editing, and enriching AI output. This demands redefining authenticity and originality: Can a text born from human ideas, structured by AI, and refined by human hands be deemed authentic? Answering this requires moving beyond traditional binaries toward transparent standards that acknowledge hybrid creation, including disclosure mechanisms.
 
4. Conclusion
The future of Persian literature hinges on achieving an intelligent equilibrium between harnessing the unprecedented capabilities of AI technology and vigilantly safeguarding the core of human creativity. This imperative balance cannot be left to organic development but demands proactive and concerted institutional action. First, there is a critical need for redefining evaluation standards within the literary field; publishers, critics, and award committees must collaboratively establish transparent criteria for disclosing AI use and forge a new, inclusive definition of “originality” that encompasses hybrid human-AI creations. Concurrently, educational transformation is essential. Literature and creative writing curricula must evolve beyond traditional forms to foster critical algorithmic literacy, empowering the next generation of writers not merely to generate content with AI, but to expertly evaluate, astutely edit, and creatively reinvent its outputs while preserving their unique authorial voice. In this new paradigm, teaching the sophisticated skill of “smart editing” becomes more crucial than teaching simple AI production. Furthermore, developing robust ethical-legal frameworks is paramount to navigate the uncharted territory of intellectual property and copyright, requiring industry bodies to collaborate with legal experts to clearly define boundaries of collaboration and ownership.
The ultimate goal of this multifaceted approach is twofold: to harness AI’s democratizing potential for including marginalized voices and lowering barriers to entry, while simultaneously preventing a descent into stylistic homogenization, qualitative decline, and the erosion of professional writers’ livelihoods. Without this smart governance, the publishing market risks being flooded with mass-produced, mediocre, and click-driven content where depth and authenticity are sacrificed for quantity and speed. It must be emphasized that AI tools like ChatGPT are ultimately just instruments, serving human will and creativity.