Assignment #4 The use of AI in publications
AI definitions and types in general
There are various definitions of Artificial Intelligence. The definition of “intelligence” varies depending on the source. Oxford dictionaries define it as “the ability to acquire and apply knowledge and skills,” while Wiktionary defines it as “the capacity of mind, especially to understand principles, truths, facts, or meanings; the ability to learn and comprehend.” (El Hadi, n.d.) Artificial intelligence (AI) is the creation of algorithms and computer programs that can learn from and make predictions or judgments based on real-world data, thus simulating human intellect. Furthermore, AI systems may be trained to recognize data patterns, make predictions, and learn from experience. AI can complete jobs that require human-like reasoning and decision-making (Giglio & da Costa, 2023).
Various types of artificial intelligence (AI) consider the variations of real-world issues (Sarker, 2022)
· Analytical AI identifies, interprets, and communicates important patterns in data. Thus, Analytical AI seeks to identify new insights, patterns, and correlations or dependencies in data, as well as to assist in data-driven decision-making.
· Functional AI acts rather than giving recommendations. For example, a functional AI model could be useful in robotics and IoT applications that require fast action.
· Interactive AI provides efficient and interactive communication automation, a common practice in daily life, especially in business. To create an interactive AI model, it is possible to use techniques like machine learning, pattern mining, reasoning, and heuristic search.
· Textual AI services include text analytics, natural language processing, speech-to-text conversion, machine translation, and content development for organizations.
· Visual AI can recognize, classify, sort, and turn photos and videos into insights. Computer vision and augmented reality use this AI.
The use of AI in some fields
AI has a wide scope in various fields, such as education, entertainment, customer service, and healthcare (Verma, 2018). In education, AI can automate grading and adapt educational software to student needs (Chavan et al., 2004). In entertainment, AI can compose music, generate stories, and even bring deceased stars back to life (Deepa & Aruna Devi, 2011). For customer service, AI is used for accurate calculations and natural language processing to provide efficient services (Deepa & Aruna Devi, 2011). In healthcare, AI is utilized for diagnosis and improving medical processes (Ramesh et al., 2004). Overall, AI technologies like expert systems, neural networks, and natural language processing are making significant impacts across different sectors (Pannu, 2008).
The use of AI in scientific publications
AI can improve data processing, task automation, and personalized experiences in research. AI utilization in research and scientific writing may increase bias reinforcement, data privacy concerns, data mistakes, and impair critical thinking due to overreliance (Bahammam, 2023). To ensure the ethical use of AI in research and scientific writing, criteria must be developed.
AI is increasingly being utilized in scientific writing to assist with various aspects of the process, such as searching for relevant papers, correcting grammatical errors, and improving writing style. AI tools like Elicit, ResearchRabbit, Bing, Google Bard, and ChatGPT can enhance the quality of scientific writing, increasing the chances of publication in major journals (Bahammam, 2023; Giglio & da Costa, 2023). However, it is essential to note that AI should be used as a tool to improve writing and not as a replacement for human authors. AI can help with tasks like summarizing references, correcting errors, and restructuring text, benefiting non-native English-speaking scientists in particular.
Ethics and rules for using AI in publications
Authorship in scientific papers entails significant contributions to the conception, design, analysis, interpretation, and writing, as well as approving the final manuscript. Additionally, authors must take public responsibility for the content. Since ChatGPT cannot fulfil these criteria, it cannot be considered a coauthor in a scientific paper (Giglio & da Costa, 2023).
Journals have varying policies regarding the use of generative AI for scientific writing, with some requiring explicit permission or detailed disclosure in manuscripts. Regardless of these policies, transparency and maintaining content integrity are universally emphasized. However, there is a potential stigma surrounding AI-assisted manuscripts, which may lead some authors to avoid disclosing AI assistance. To mitigate this, education initiatives are essential for authors, reviewers, and editors to understand the ethical use of AI tools in academic writing and emphasize their role in augmenting human creativity rather than replacing it. Publishers like Dove Press explicitly require authors to acknowledge any AI tool assistance and uphold ethical standards in their writing process (Bahammam, 2023).
AI-generated texts present ethical concerns, including the risk of plagiarism. While plagiarism detectors can identify copied text, automatic rephrasing may evade detection. This technology could lead to an increase in papers lacking substantial contributions, exacerbating the decline in scientific breakthroughs (Giglio & da Costa, 2023).
AI-based systems and platforms for publications
The utilization of artificial intelligence (AI) tools, including large language models (LLMs) like ChatGPT, Google Bard, Bing AI, and others, is on the rise in research publications (Bahammam, 2023). The Midjourney AI, DALL-E 2, and Deep Dream Generator AI are all OpenAI systems that can make realistic pictures and art from natural language descriptions (Bahammam, 2023).
Two AI programs, Research Rabbit (https://www.researchrabbit.ai/) and Elicit (www.elicit.org), can be beneficial for suggesting references that may not yet be known to scientists. These programs employ a self-learning approach based on the content, keywords, and citations of retrieved papers to go “beyond the horizon” and find other relevant papers (Giglio & da Costa, 2023).
The latest hot issues about AI-assisted publications that are circulating in social media currently
One hot issue regarding AI-assisted publications revolves around the ethical implications and challenges associated with AI-generated content and authorship. As AI tools become more sophisticated, they increasingly contribute to various stages of the research and publication process, raising questions about their role and recognition in scholarly work. AI’s capacity to generate text raises concerns about plagiarism and originality. While AI can aid in content creation and summarization, there is a risk that generated text may inadvertently replicate existing material without proper attribution. Ensuring the originality and integrity of AI-assisted publications presents a significant challenge for researchers and publishers.
AI models trained on biased or incomplete datasets may produce biased outputs, perpetuating existing inequalities and misconceptions. Addressing bias in AI-assisted publications requires careful attention to dataset selection, algorithm design, and evaluation methods to mitigate potential biases and promote fairness in scholarly communication.
Illustration of how to use AI in creating the introduction part of a scientific paper
A scientific paper introduction can be written faster and better with AI. Define the study topic, acquire background information, select an AI tool, enter the necessary information, review and revise the text, incorporate human knowledge, cite important literature, finalize and modify the introduction, and seek feedback from colleagues, mentors, or collaborators to apply AI. This procedure promotes clarity, coherence, and accuracy, supporting the research paper’s goals and structure. Citing relevant literature and following journal criteria is crucial.
References
Bahammam, A. S. (2023). Balancing Innovation and Integrity: The Role of AI in Research and Scientific Writing. In Nature and Science of Sleep (Vol. 15, pp. 1153–1156). Dove Medical Press Ltd. https://doi.org/10.2147/NSS.S455765
Chavan, S., Shah, K., Dave, N., Mukherjee, S., Abraham, A., & Sanyal, S. (2004). Adaptive Neuro-Fuzzy Intrusion Detection Systems.
Deepa, S., & Aruna Devi, B. (2011). A survey on artificial intelligence approaches for medical image classification. Indian Journal of Science and Technology, 4(11). http://www.indjst.org
El Hadi, M. M. (n.d.). Artificial Intelligence Background, Definitions, Challenges and Benefits.
Giglio, A. Del, & da Costa, M. U. P. (2023). The use of artificial intelligence to improve the scientific writing of non-native english speakers. Revista Da Associacao Medica Brasileira, 69(9). https://doi.org/10.1590/1806-9282.20230560
Pannu, A. (2008). Artificial Intelligence and its Application in Different Areas. In Certified International Journal of Engineering and Innovative Technology (IJEIT) (Vol. 9001, Issue 10).
Ramesh, A. N., Kambhampati, C., Monson, J. R. T., & Drew, P. J. (2004). Artificial intelligence in medicine. In Annals of the Royal College of Surgeons of England (Vol. 86, Issue 5, pp. 334–338). https://doi.org/10.1308/147870804290
Sarker, I. H. (2022). AI-Based Modeling: Techniques, Applications and Research Issues Towards Automation, Intelligent and Smart Systems. SN Computer Science, 3(2). https://doi.org/10.1007/s42979-022-01043-x
Verma, M. (2018). Artificial intelligence and its scope in different areas with special reference to the field of education. In International Journal of Advanced Educational Research 5 International Journal of Advanced Educational Research (Vol. 3). www.educationjournal.org
