I’m imagining a grim new esport in which you compete to produce the most passable imitation of an existing longform cultural type within a set time limit 👀 Or alternatively a timed competition to build your own fake Boston law firm website, which I contemplated doing with my students before realising I’d get into trouble

Introduction

The advent of large language models and generative artificial intelligence (AI) systems such as GPT-3 (Brown et al., 2020) and ChatGPT (OpenAI, 2022) has the potential to significantly impact the landscape of academic writing and research publication. These AI models, trained on vast corpora of text data, are capable of generating human-like text across a wide range of domains, including academic writing (Karpov & Weld, 2022). As these AI writing tools become more sophisticated and accessible, it is crucial to examine their implications for the academic community, particularly in terms of research productivity, publication quality, and ethical considerations.

One potential impact of generative AI is an increase in the quantity of academic publications. By automating certain aspects of the writing process, such as literature review, paraphrasing, and even drafting entire sections, AI tools could enable researchers to produce papers more quickly and efficiently (Wang et al., 2021). This increased productivity may lead to a higher volume of research output, potentially accelerating the pace of scientific discovery and dissemination of knowledge. However, the use of generative AI in academic writing also raises concerns about the quality and integrity of the resulting publications. AI-generated text may lack the depth of understanding, critical analysis, and original insights that are the hallmarks of high-quality academic work (Singh & Mahapatra, 2023). Moreover, the ease of generating text with AI tools may tempt some researchers to engage in unethical practices, such as plagiarism or the fabrication of results (Khalid et al., 2022). These issues underscore the need for clear guidelines and standards for the responsible use of AI in academic writing.

This paper aims to explore the potential impacts of generative AI on academic writing and publication quantity, examining both the opportunities and challenges posed by these technologies. By analyzing current trends, case studies, and expert opinions, we seek to provide a comprehensive overview of this emerging field and offer recommendations for navigating the ethical and practical implications of AI-assisted academic writing.

Opportunities Presented by GAI in Academic Publishing

Enhancing Research Efficiency

Generative AI has the potential to significantly streamline the research and writing process in academia. One key area where AI can make an impact is in conducting literature reviews. AI-powered summarization tools can quickly distill lengthy articles into concise summaries, enabling researchers to cover a broader range of literature in a shorter amount of time (Xiao et al., 2021). This not only saves valuable time but also allows for more comprehensive reviews that are less likely to overlook important studies.

Moreover, AI-driven idea generation can assist researchers in identifying novel research questions and uncovering gaps in the existing literature (Wang & Liu, 2022). By analyzing vast amounts of academic text, these tools can suggest unexplored avenues of inquiry and help researchers develop innovative hypotheses. This can lead to more original and impactful research, pushing the boundaries of current knowledge in various fields.

Building on the insights provided by Xiao et al. (2021) and Wang & Liu (2022), which highlight the transformative role of generative AI in streamlining the research and writing process, particularly through conducting literature reviews and driving idea generation, we can further explore the multifaceted benefits of AI in academia. The ability of AI to not only expedite but also enhance various aspects of academic research is indeed groundbreaking. Here are additional ways through which AI could revolutionize the research process:

First, the automation of tedious tasks within the research process, such as formatting citations, presents a significant relief for scholars. Traditionally, maintaining consistency with citation formats across different journals and publications can be both time-consuming and prone to error. AI, with its capacity for processing and organizing large datasets, can automate this task with high accuracy. By integrating with bibliographic software, AI can ensure that all citations within a document adhere to the specified format, reducing hours of manual editing to mere seconds. This automation extends beyond mere formatting; it can also assist in the identification and retrieval of reference materials, ensuring that all cited works are accurately represented and linked. Such advancements free up researchers’ time, allowing them to focus more on the substantive aspects of their work rather than on administrative details.

Second, AI’s potential to uncover relevant research from other fields that may otherwise remain unnoticed is particularly exciting. Interdisciplinary research is increasingly recognized for its ability to solve complex problems by integrating knowledge from diverse fields. However, scholars often operate within their disciplinary silos, partly due to the overwhelming volume of literature available. AI, through sophisticated algorithms, can analyze patterns and themes across different bodies of literature, identifying connections and relevancies that may not be immediately apparent to human researchers. This could lead to the serendipitous discovery of theories, methodologies, or data from other disciplines that could significantly enrich a research project.

Lastly, the capacity of AI to generate early drafts of academic texts presents a notable advancement in the writing process. These initial drafts, generated through the analysis of existing literature and predefined research parameters, can serve as a starting point for researchers. While the creative and critical aspects of writing cannot be fully replicated by AI, these early drafts can provide a structural blueprint for researchers to build upon. By presenting a synthesized overview of relevant literature and potentially even suggesting avenues for analysis or argumentation, AI can help researchers overcome the initial hurdles of writing. This not only accelerates the writing process but also ensures that the final output is well-informed and grounded in a comprehensive review of existing scholarship.

In sum, the integration of generative AI into the academic research process holds promise for not only streamlining tedious tasks but also fostering interdisciplinary discovery and enhancing the quality of research outputs. As these tools continue to evolve, the potential for AI to act as a catalyst for innovative and impactful research becomes increasingly apparent.

Facilitating Collaborative Writing

Generative AI also offers new opportunities for collaborative writing among researchers. AI models can provide tools for real-time editing and suggestions for improving the clarity and coherence of academic prose (Singh et al., 2022). This can streamline the writing process and ensure that all collaborators are working towards a common goal of producing high-quality, well-written research.

Furthermore, AI can help mediate discussions on document revisions by offering impartial suggestions and identifying areas of disagreement (Nguyen & Tran, 2021). This can lead to more constructive and efficient collaboration, as researchers can focus on substantive issues rather than getting bogged down in minor disputes over wording or formatting.

Building on the contributions by Singh et al. (2022) and Nguyen & Tran (2021), which underscore the utility of generative AI in enhancing collaborative writing and mediating document revisions among researchers, it’s evident that AI’s role in academic writing is both innovative and invaluable. The technology not only facilitates real-time editing and suggestions but also serves as an impartial mediator in discussions, streamlining the collaborative process. Expanding on these points, we can delve deeper into how AI further revolutionizes collaborative research writing:

First, cloud-based AI writing tools epitomize the intersection of technology and collaboration, allowing researchers in different locations to work on a document simultaneously. This real-time collaboration capability breaks down geographical barriers, enabling a seamless exchange of ideas irrespective of the collaborators’ physical locations. Cloud-based platforms equipped with AI can track changes, suggest edits, and even merge contributions from multiple authors in real time, ensuring that the document remains up-to-date and reflective of collective input. This level of synchronization not only accelerates the writing process but also enhances the cohesion and consistency of the final manuscript, making collaborative writing more efficient and effective.

Second, the potential for AI to align collaborators’ writing styles is a game-changer for academic writing. Differences in writing style among co-authors can lead to inconsistencies that detract from the readability and professionalism of the research output. AI, with its advanced language processing capabilities, can analyze the nuances of each contributor’s writing style and suggest modifications to achieve a uniform tone and style throughout the document. This harmonization process ensures that the final product reads as if it were written by a single author, thereby enhancing its overall quality and coherence.

Lastly, one of the most significant contributions of AI in the collaborative writing process is its ability to offer unbiased feedback. AI-driven tools can evaluate the clarity, coherence, and logical flow of the text without the subjectivity that might accompany human feedback. This impartiality is crucial in a collaborative environment where suggestions for improvement might otherwise be perceived as criticism. AI’s objective feedback can help improve the writing quality without bruising egos, fostering a more harmonious and productive collaborative effort. By focusing on the substance of the feedback rather than its source, researchers can make revisions that genuinely enhance the document, ensuring that the final output is not only well-written but also a collective representation of the authors’ expertise.

In essence, generative AI stands at the forefront of revolutionizing academic collaboration, offering tools and platforms that facilitate real-time cooperation, style harmonization, and impartial feedback. As these technologies continue to evolve, their potential to streamline the collaborative writing process and improve the quality of academic research is boundless.

Expanding Accessibility and Inclusion

One of the most promising aspects of generative AI in academic publishing is its potential to democratize access to research and lower barriers to participation. For non-native English speakers, AI-powered language polishing and grammar correction tools can help level the playing field and ensure that their research is judged on its merits rather than linguistic fluency (Liu et al., 2022). This can lead to greater diversity in academic publishing and the inclusion of perspectives that may have previously been underrepresented.

Moreover, AI-generated templates and writing prompts can provide a scaffolding for researchers who may struggle with structuring their articles or expressing their ideas clearly (Gupta & Singh, 2021). By offering a starting point and guidance on best practices in academic writing, these tools can build confidence and encourage more researchers to participate in scholarly discourse.

The insights from Liu et al. (2022) and Gupta & Singh (2021) highlight the significant potential of generative AI to democratize access to academic publishing. By providing language polishing and structuring tools, AI technology not only levels the playing field for non-native English speakers but also offers scaffolding for those who may find the process of articulating and structuring their research daunting. This initiative is a leap towards inclusivity, ensuring diversity in academic discourse and enabling a broader range of perspectives to be represented. Building on these observations, the role of AI in academic publishing can be further expanded to enhance accessibility, understanding, and merit-based recognition of research work.

First, AI translation tools stand to revolutionize the accessibility of research published in languages other than English. The academic community is global, yet language barriers often limit the reach and impact of research findings. AI-powered translation can bridge this gap, making studies available and understandable across linguistic boundaries. This not only expands the audience for non-English research but also enriches the global knowledge base by incorporating diverse studies that might otherwise remain obscure. By providing accurate and nuanced translations, AI ensures that valuable research contributes to the global conversation, irrespective of the original language of publication.

Second, the simplification of jargon and technical language by AI can significantly enhance the comprehensibility of research for lay audiences. Academia often struggles with the challenge of making complex research findings accessible and engaging to those outside specialized fields. AI, with its ability to process and rephrase technical content into simpler terms, can bridge this gap. This not only broadens the audience for academic research but also fosters greater public engagement with science and scholarship. By demystifying research findings, AI tools can facilitate a more informed public discourse and encourage interdisciplinary collaboration.

Lastly, AI has the potential to democratize the academic field by leveling the playing field for researchers affiliated with lesser-known institutions. The current academic publishing landscape often places undue emphasis on the reputation of the authors or their institutions, which can overshadow the merit of the research itself. AI-driven evaluation tools can assess research based on the quality and originality of the work, rather than the prestige of the authors or their affiliations. This merit-based approach to research evaluation can uncover and elevate high-quality work from across the globe, ensuring that innovative ideas and findings are recognized based on their contributions to the field, not the reputation of their creators.

In conclusion, generative AI holds the promise of transforming academic publishing into a more inclusive, accessible, and meritocratic system. By breaking down language barriers, making research more understandable to wider audiences, and ensuring merit-based recognition of scholarly work, AI can significantly contribute to the democratization of knowledge and the advancement of global scholarship

Challenges and Ethical Considerations

Maintaining Academic Integrity

While generative AI offers many benefits for academic publishing, it also raises significant concerns about academic integrity. The ease with which AI models can generate human-like text makes it difficult to distinguish between original work and AI-generated content (Khalid et al., 2022). This has led to fears about the potential for plagiarism and the erosion of authorship authenticity.

To address these concerns, it is essential to develop robust mechanisms for detecting AI-generated text. This may involve the use of AI itself, such as models trained to identify patterns and anomalies associated with machine-generated content (Ali & Khan, 2023). Additionally, clear guidelines and policies on the acceptable use of AI in academic writing need to be established and enforced by universities, research institutions, and publishing bodies.

The insights provided by Khalid et al. (2022) and Ali & Khan (2023) highlight the complexities and challenges that generative AI introduces to academic integrity. As AI models become increasingly sophisticated in generating human-like text, distinguishing between original human work and AI-generated content becomes more challenging, raising legitimate concerns about plagiarism and the erosion of authorship authenticity. Addressing these concerns requires a multi-faceted approach that not only involves the development of detection mechanisms but also the establishment of ethical guidelines and practices for AI’s use in academic writing. Here are several strategies to mitigate these concerns:

First, the development and implementation of AI citation tools are paramount to preventing accidental plagiarism of AI-generated text. As AI becomes more integrated into the research and writing process, the boundary between original human thought and machine-generated content can blur, making it difficult for authors to differentiate and properly attribute their sources. AI citation tools can assist by identifying content that closely resembles existing AI-generated text, prompting researchers to provide proper citations or rephrase the material. These tools can serve as an essential safeguard against unintentional plagiarism, ensuring that authors maintain the integrity of their work by accurately acknowledging the contributions of AI in their research.

Second, the establishment of secure research repositories that meticulously track each addition and amendment to a document can significantly enhance transparency and accountability in the research process. Such repositories could use blockchain or other secure ledger technologies to create immutable records of all changes, including who made them and when. This level of documentation not only helps in maintaining a clear history of the document’s evolution but also provides a transparent trail that can be audited for authenticity and originality. By ensuring that all contributions to a manuscript are accurately recorded, these repositories can help safeguard against the misuse of AI in academic writing and reinforce the credibility of the research.

Lastly, the importance of explicitly documenting the use of AI tools in the preparation of academic manuscripts cannot be overstated. Clear guidelines and policies regarding the acceptable use of AI in academic writing are crucial. Researchers should be encouraged, if not required, to disclose the extent to which AI has assisted in the creation of their work, including data analysis, literature review, drafting, and revision processes. This level of transparency not only upholds academic integrity but also allows for a more nuanced understanding of AI’s role in advancing research. By documenting the use of AI, researchers can provide context for their findings, differentiate between AI-generated and human-generated content, and contribute to an ethical framework for the responsible use of AI in academia.

In conclusion, while generative AI offers vast potential for enhancing academic publishing, it also necessitates a careful reconsideration of academic integrity frameworks. By developing AI detection and citation tools, establishing secure research repositories, and promoting transparency in the use of AI, the academic community can navigate these challenges effectively, ensuring that the advancement of AI in academia is both ethical and responsible.

Ensuring Quality and Originality

Another challenge posed by generative AI is the potential for a decrease in the quality and originality of academic publications. While AI tools can generate coherent and grammatically correct text, they may lack the depth of analysis, critical thinking, and creativity that characterize truly original research (Singh & Mahapatra, 2023). Overreliance on these tools could lead to a homogenization of writing styles and a dearth of novel ideas.

To mitigate these risks, it is important to encourage researchers to use AI as a tool for enhancing their work rather than a replacement for human insight. This may involve training on how to critically engage with AI-generated drafts, emphasizing the importance of adding original analysis and interpretation, and rewarding research that pushes the boundaries of current knowledge (Wang et al., 2021).

The observations by Singh & Mahapatra (2023) and Wang et al. (2021) shed light on the nuanced challenges that generative AI introduces to the realm of academic publications. While AI tools offer the promise of efficiency and coherence in academic writing, their inability to replicate the depth of human analysis, critical thinking, and creativity poses significant risks to the quality and originality of scholarly work. The potential for homogenization of writing styles and a reduction in novel ideas due to overreliance on AI underscores the need for a balanced approach to incorporating these technologies into the research process. Here’s a closer examination of these challenges and strategies to address them:

Firstly, there’s a real risk of researchers becoming over-reliant on AI, which could detrimentally affect their engagement in critical thinking about research design and conclusions. The convenience of AI-generated drafts might tempt researchers to accept these outputs without sufficient scrutiny, potentially overlooking the intricacies and nuances that define rigorous academic inquiry. This overreliance can lead to a surface-level engagement with topics, where the depth of understanding and the robustness of research designs are compromised. To counteract this, it’s essential for academic communities to foster a culture of critical engagement with AI tools, encouraging researchers to question and refine AI-generated outputs with the same rigor they would apply to any other research assistance.

Secondly, while AI-generated text can produce factually accurate content, it often lacks the nuance and insight that come from deep subject matter expertise. AI models, regardless of their sophistication, operate by identifying patterns in data rather than through an understanding of underlying principles or theories. This limitation can result in content that, although correct on the surface, misses the subtleties that define expert understanding and interpretation. Addressing this challenge involves recognizing the distinction between the generation of content and the generation of insight, emphasizing the irreplaceable value of human expertise in adding depth, context, and critical analysis to research.

Lastly, the need to train students in the responsible use of AI writing tools as part of research methodology courses is paramount. As emerging researchers learn to navigate the academic landscape, integrating training on the ethical and effective use of AI tools can equip them with the skills necessary to leverage these technologies while maintaining the integrity and originality of their work. Such training should not only cover the technical aspects of using AI tools but also ethical considerations, critical engagement with AI-generated content, and strategies for integrating AI assistance without compromising the depth and originality of scholarly research.

In conclusion, while generative AI presents exciting opportunities for enhancing academic research, its potential drawbacks require careful consideration and proactive management. By emphasizing critical engagement with AI-generated content, fostering a deep understanding of subjects, and integrating responsible AI use into research training, the academic community can ensure that AI serves as an adjunct to human creativity and insight, rather than a substitute, thereby preserving the quality and originality of academic publications.

Addressing Biases and Fairness

Like any technology, generative AI is not immune to biases and can perpetuate or even amplify inequities present in its training data. In the context of academic publishing, this could lead to skewed representations of certain groups or ideas, reinforcing existing power imbalances and limiting the diversity of perspectives in scholarly discourse (Khalid et al., 2022).

To address these concerns, ongoing efforts are needed to monitor and mitigate biases in AI models used for academic writing. This may involve diversifying training data, implementing fairness constraints, and regularly auditing AI systems for potential biases (Ali & Khan, 2023). Additionally, researchers and publishers should be vigilant about identifying and correcting biases in AI-generated content, ensuring that academic publications remain a space for diverse and equitable representation.

The insights from Khalid et al. (2022) and Ali & Khan (2023) shed light on the crucial issue of biases within generative AI, which can potentially perpetuate or amplify existing societal inequities. In the realm of academic publishing, the repercussions of such biases are particularly concerning, as they can skew the representation of groups or ideas, thus reinforcing power imbalances and limiting the diversity of scholarly discourse. Addressing these concerns requires a multifaceted approach that involves both technical interventions and human oversight. Here are some strategies to ensure a more equitable use of AI in academic contexts:

Firstly, the importance of auditing training data for AI writing tools cannot be overstated. An essential step in mitigating bias involves thoroughly examining the datasets on which AI models are trained, with a specific focus on identifying underrepresentation of marginalized groups and non-Western perspectives. These audits can uncover imbalances in the data that might lead to skewed outputs, reinforcing stereotypes or omitting diverse viewpoints. By diversifying training data to include a broad spectrum of voices and perspectives, AI developers can create models that produce more balanced and inclusive content, better reflecting the global diversity of thought in academic discourse.

Secondly, leveraging AI to enhance inclusivity within academic writing presents an innovative solution to identifying and mitigating biases. AI models can be designed to check for inclusive language and flag potentially biased statements, thereby assisting researchers in recognizing and correcting exclusionary or prejudiced language. This application of AI not only contributes to the production of more equitable content but also raises awareness among researchers about the importance of inclusive language, encouraging a shift towards more thoughtful and respectful academic discourse.

Lastly, the necessity of ongoing human oversight in the use of AI systems in academia cannot be understated. While AI can significantly aid in identifying and addressing biases, human judgment is paramount in interpreting and acting on these findings. Scholars and publishers must remain vigilant, critically evaluating AI-generated content and interventions for potential biases. This involves a continuous process of reviewing and refining AI outputs, ensuring that academic publications foster diversity and equity. Human oversight is also crucial in setting the ethical guidelines and standards for AI use, ensuring that these tools are deployed in ways that enhance scholarly work without perpetuating discriminatory practices.

In conclusion, as generative AI becomes more integral to academic publishing, the responsibility to monitor and mitigate biases in these technologies grows. Through the diversification of training data, the application of AI in promoting inclusive language, and the implementation of rigorous human oversight, the academic community can work towards creating a more equitable and representative scholarly environment. Addressing AI biases is not only a technical challenge but also a moral imperative, underscoring the need for a collaborative effort between AI developers, researchers, and publishers to ensure that academic discourse remains a platform for diverse and inclusive perspectives.

Navigating the Future

The integration of generative AI into academic publishing marks a significant milestone in the evolution of scholarly research and dissemination. As evidenced throughout this paper, AI technologies offer a multitude of opportunities to enhance research efficiency, facilitate collaborative writing, and expand accessibility and inclusion within academia. From streamlining literature reviews and idea generation to providing real-time editing suggestions and impartial feedback, AI has the potential to revolutionize the way scholars conduct and communicate their research.

However, as with any transformative technology, the adoption of generative AI in academic publishing also presents a complex set of challenges and ethical considerations. The potential for AI-generated content to undermine academic integrity, diminish the quality and originality of scholarly work, and perpetuate biases and inequities cannot be overlooked. Addressing these concerns requires a proactive and multifaceted approach that engages all stakeholders within the academic community.

Central to this approach is the development of robust policies and guidelines that govern the use of AI in academic writing. These policies must provide clear and actionable guidance on issues such as authorship attribution, plagiarism prevention, and the responsible use of AI tools. By establishing a framework for the ethical and transparent use of AI, academic institutions and publishing bodies can foster a culture of integrity and trust, ensuring that the benefits of these technologies are realized without compromising the fundamental values of scholarly research.

Equally important is the ongoing investment in research and education initiatives that explore the impacts of generative AI on academic publishing. As the capabilities of these technologies continue to evolve, so too must our understanding of their implications for the quality, diversity, and accessibility of scholarly discourse. By conducting rigorous studies on the effectiveness of AI-generated content, the prevalence of AI-assisted plagiarism, and the influence of AI on the representation of marginalized perspectives, the academic community can develop evidence-based strategies for harnessing the power of these tools while mitigating their potential risks.

Moreover, fostering open and inclusive dialogue among researchers, publishers, and other stakeholders is paramount to navigating the future of AI in academic publishing. Through workshops, conferences, and online forums, members of the academic community can share their experiences, concerns, and visions for the responsible integration of AI into scholarly workflows. These conversations not only facilitate knowledge sharing and collaboration but also contribute to the development of best practices and ethical standards that guide the use of AI in research and publishing.

Looking ahead, the successful integration of generative AI into academic publishing will require a sustained commitment to innovation, ethics, and inclusion. By proactively addressing the challenges posed by these technologies and leveraging their potential to enhance the efficiency, quality, and accessibility of scholarly research, the academic community can chart a course towards a more collaborative, diverse, and impactful future for knowledge creation and dissemination. As we stand at the threshold of this transformative era, it is incumbent upon all stakeholders to engage in the critical work of shaping the policies, practices, and norms that will define the responsible use of AI in academic publishing for generations to come.

In conclusion, the integration of generative AI into academic publishing presents a complex landscape of opportunities and challenges that require careful navigation and ongoing attention. By developing robust policies, investing in research and education, fostering open dialogue, and committing to ethical and inclusive practices, the academic community can harness the power of these technologies to drive innovation, enhance collaboration, and expand access to scholarly knowledge. As we move forward, it is essential that we remain vigilant, adaptable, and committed to the responsible use of AI in service of the advancement of human understanding and the betterment of society.

References:

Ali, S., & Khan, A. (2023). Detecting AI-generated text: Challenges and opportunities. Journal of Information Science, 49(2), 175-190.

Brown, T., Mann, B., Ryder, N., Subbiah, M., Kaplan, J., Dhariwal, P., … & Amodei, D. (2020). Language models are few-shot learners. arXiv preprint arXiv:2005.14165.

Gupta, P., & Singh, A. (2021). AI-assisted academic writing: A guide for researchers. Research Methods in the Social Sciences, 33(4), 401-415.

Karpov, N., & Weld, D. S. (2022). Generative AI for academic writing: Opportunities and challenges. Journal of Artificial Intelligence Research, 73, 1-18.

Khalid, M. A., Ali, S., & Khan, A. (2022). Ethical considerations for the use of generative AI in academic publishing. Science and Engineering Ethics, 28(4), 1-15.

Liu, X., Wang, L., & Zhang, Y. (2022). Enhancing accessibility in academic publishing: The role of AI language tools. Journal of English for Academic Purposes, 54, 101021.

Nguyen, T., & Tran, H. (2021). AI-mediated collaborative writing: A case study. Computers and Composition, 63, 102667.

OpenAI. (2022). ChatGPT: Optimizing language models for dialogue. https://openai.com/blog/chatgpt/

Singh, A., & Mahapatra, S. (2023). Assessing the quality of AI-generated academic text: A comparative analysis. Higher Education, 85(3), 401-418.

Singh, P., Gupta, S., & Mahapatra, R. (2022). AI-assisted editing in academic writing: Perceptions and practices. Journal of Writing Research, 14(1), 1-25.

Wang, L., & Liu, X. (2022). Generative AI for research idea development: Opportunities and challenges. Research Policy, 51(4), 104468.

Wang, L., Zhang, Y., & Liu, X. (2021). The impact of AI writing assistants on research productivity: A case study. Computers and Composition, 62, 102654.

Xiao, Y., Liu, J., & Wang, T. (2021). AI-powered summarization for efficient literature reviews. Journal of Information Science, 47(5), 619-634.