New Study Explores AI Chatbots in Veterinary Neurology
The use of AI chatbots has gained significant popularity in various fields, and now a new study is delving into their potential in generating scientific content in the field of veterinary neurology. According to TechXplore, the study evaluated abstracts and introductory sections of both original research papers and those generated by ChatGPT, an AI chatbot, using AI output detectors and plagiarism detectors.
To determine the originality and reliability of the content, the study also involved the opinions of 13 Board-Certified neurologists. The researchers aimed to understand the nuanced use of AI tools, highlighting the importance of proper education and interdisciplinary collaboration to establish responsible guidelines for AI use in academia.
Lead author Samira Abani from the University of Veterinary Medicine Hannover stressed the need for understanding how AI tools can either pose a threat to scientific integrity or assist researchers, depending on their usage. Abani recommended integrating education on both the proper use and potential misuse of AI-based tools in academia as part of good scientific practice.
Corresponding author Jasmin Neßler also emphasized the need for interdisciplinary scientific collaboration to establish clear guidelines for responsible AI use. Neßler stated that embracing AI’s potential for the benefit of society instead of banning AI tools is the most effective approach.
The study unveiled challenges in distinguishing between ChatGPT-generated and human-written abstracts, especially when the subject matter was less familiar. Only four out of 13 reviewers accurately identified the AI-generated text in such cases. However, familiarity with the topic improved accuracy, with seven out of 13 reviewers correctly identifying the AI-written abstract.
The inclusion of the introduction and references significantly aided reviewers in differentiating between human-written and AI-generated content, leading to a notable increase in their performance. The study raises concerns about the need for caution and proper education in the use of AI-based tools to maintain the integrity and transparency of published literature.
The increasing influence of AI in academic writing calls for interdisciplinary collaboration to establish clear guidelines for responsible AI application in scientific practice. The authors suggested integrating education on the proper use and potential misuse of AI-based tools in academia for both pre- and post-graduate students.
The study’s findings have been published in the journal Frontiers in Veterinary Science. It highlights the potential of AI-based tools in scientific writing but also underscores the importance of caution and proper education in their use.
