Generative artificial intelligence and its impact on writing

• Ano de defesa: 2024
• Autor(a) principal: Santos, Allessandra Elisabeth dos
• Orientador(a): Silva, Paulo Roberto Boa Sorte
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: eng
• Programa de Pós-Graduação: Pós-Graduação em Educação
• Palavras-chave em Português: Educação; Inteligência artificial generativa; ChatGPT; Modelos de linguagem de grande escala; Escrita
• Palavras-chave em Inglês: Education; Generative artificial intelligence; Large language models; Writing
• Área do conhecimento CNPq: CIENCIAS HUMANAS::EDUCACAO
• Link de acesso: https://ri.ufs.br/jspui/handle/riufs/20805

AI: word of 2023. The abbreviation of Artificial Intelligence (AI) was named the word of the year by Collins Dictionary, reflecting its growing prominence in everyday life. Although AI has long permeated various aspects of human activity, our society only began recognizing its ubiquitous presence after ChatGPT was released on November 30, 2022. ChatGPT, an AI chatbot that produces human-like responses, exemplifies Generative Artificial Intelligence (GenAI) technologies now widely accessible, provoking discussions and raising ethical concerns across social, political, economic, and academic contexts. In education, technologies undeniably reshaped writing, a social practice grounded in evolving literacy perspectives. Recognizing these transformations, the central question of this study is: How does the integration of GenAI technologies, like Large Language Models (LLMs), redefine the concept of writing in educational contexts? To address this question, this research aims to apprehend the concept of the phenomenon of writing impacted by the adoption of GenAI in societal contexts in the field of education. Additionally, the specific objectives are to examine the approach to writing conveyed on internet-based platforms, particularly after the launch of ChatGPT, and to explore educational, pedagogical, and ethical issues related to the impact of AI-text generators, especially ChatGPT, raised on these platforms. Considering that the study of the phenomenon of writing cannot be done in isolation from our society, knowledge, and culture, it is crucial to adopt an epistemological and methodological approach that rejects oversimplifying realities. Guided by Complex Thinking (Morin, 2015), this qualitative, documentary, and bibliographic study analyzed 60 documents — including digital newspaper articles, social media posts, YouTube videos, podcasts, and websites — using Analytical Entwining integrated with GenAI. Findings revealed eight theoretical constructs. The key terms — revolution, integrity, and transformation — were extracted from the first three constructs and address the first specific objective of this research. While the key terms — enhancement, support, quality, authenticity, equity, and integration — extracted from five theoretical constructs, address the second objective. By situating writing as an evolving social practice that interacts reciprocally with GenAI, the study suggests that GenAI broadens writing as a social technology, reinforcing the need for adaptive strategies that ensure AI augments rather than replaces human agency. The integration of GenAI in writing, analyzed through the principles of Complex Thinking, reveals a complex interplay between technology and human agency, with both elements in constant adaptation, influencing each other systemically, holographically, recursively, and dialogically, ensuring that the writing process remains dynamic and grounded in humanistic values. Recognizing writing as part of this evolving system, this research introduces the concept of writing as an ecosystem, in which technology and human elements are interdependent and mutually enhancing, contributing to a more collaborative and adaptive approach to writing instruction.