Artificial intelligence—The next frontier of scientific publications?

Abstract

The science and art of publishing scholarly research has a long and robust history. Our evidence-based knowledge and practice has been created through the rigorous scientific process and human endeavour. The dissemination of research is a rapidly evolving process as technology is developed, debated, adopted, and embraced. We are in a research environment surrounded by digital tools and platforms that we now take for granted. The last decade has seen the emergence of professional academic social networks such as Research Gate and Academia, scholarly journals moving to online and open access formats (Hosseini et al., 2023). There has been rapid uptake of researchers using social media platforms to self-promote and market their work to targeted audiences beyond the usual readership of journals. The impact of COVID-19 has also influenced in the increasing usage of digital platforms instead of the more traditional approaches to dissemination (e.g., conference) (Ramos, 2021). We are also consumers of digital tools to enhance research capacity and analysis in systematic reviews and qualitative and quantitative methodologies (Marshall & Wallace, 2019). All however still function in the context of the accepted research process. The next major frontier in research processes and dissemination has already arrived without us almost knowing. Artificial intelligence (AI) or natural language processing (NLP) systems have the potential to be utilised in many forms of analysis and writing, including scientific and scholarly publications (Heaven, 2022). AI is a self-learning tool and simply draws on the current knowledge/content available online with variable combinations of accuracy, errors, and biases. Is this the best thing that has ever happened to research and publications, or does it present as one of the greatest risks to research integrity and ethical processes we have experienced? This is a question that will be answered over time as we come to understand the potential strengths and threat of AI. However, like other technological advances that all have strengths and challenges, AI is here to stay and will find its place in both high quality and sadly in more dubious research outcomes. It is not unreasonable to expect that AI will be used, if not already, to quickly generate literature reviews and introductions, methodologies, perform data analysis and discussions. The quality of the text will be directly dependent upon what is asked to be generated and the availability of online data. For those who have used ChatGPT, it quickly becomes evident that the quality of the output being produced is dependent on the maturity of the topic in terms of content, debate, and scholarship. It is early days but with the exponential development of AI, it will soon be a tool of trade for researchers who need to maximise time and funds for scientific work. What is important as we move forward is how AI is recognised as an asset and a tool to facilitate high quality research and assist in publishing outcomes in a more timely manner. Inevitably, there needs to be a serious focus on the protection of research and publication integrity resulting from unethical use of AI. Plagiarism and similarity checking tools will soon be able to detect unacknowledged AI created materials and as now occurs human authors will still need to accept the responsibility for the accuracy and integrity of the text being submitted. It has been suggested that researchers should acknowledge AI as a contributor to research (Hosseini et al., 2023), and ChatGPT has already be given authorship attribution on journal papers (Stokel-Walker, 2023). This is interesting as co-authors need to give approval for their work to be published and take full responsibility for its content—something that ChatGPT has not yet signed up to! The use of AI is probably better reflected in methodologies acknowledging the ethical use of non-human technologies as already occurs with propriety software analysis tools. The Australian Journal of Occupational Therapy, like other scientific publications, must embrace the potential of AI. The use of AI has the power to transform research and publishing. However, it is essential to use it responsibly and maintain research integrity. While AI can support research, it should not replace human expertise and critical thinking. As with any technological advancement, we must carefully navigate the opportunities and risks to ensure that AI is utilised to advance the pursuit of knowledge and improve our understanding of the world around us. (Full disclosure: Last paragraph generated by AI.)

Ähnliche Arbeiten