Editorial: Chatbots as humanlike text generators: friend or foe?

Abstract

The field of artificial intelligence, particularly in the realm of natural language processing, has seen significant advancements with the development of chatbots like ChatGPT. These AI-driven text generators have been increasingly utilized across diverse sectors such as education, science, law, and health, offering users a novel way to access information and assistance. Despite their growing popularity, there remains a paucity of empirical research examining the real-world impact of these tools on users. Key questions persist regarding how individuals perceive the utility of chatbots, the nature of their interactions, and whether these digital entities are seen as allies or adversaries. While existing studies have employed SWOT analyses to explore the strengths and weaknesses of chatbots (e.g., Farrokhnia et al. 2023), and others have delved into their conversational dynamics and ethical implications (e.g., Loos et al., 2023), there is a notable gap in understanding the user experience and the broader societal implications of chatbot integration. The contributions in this research topic explore the nuanced relationship between users and chatbots, to uncover the factors that influence user perceptions and behaviors, and to determine whether chatbots are viewed as beneficial companions or potential threats. This research topic includes the following five contributions, covering diverse users (patients, caregivers, and clinicians, physicians, nurses, regulated mental health professionals, educators, students), various chatbot user domains (health, education and law), and countries (Cannada, Germany, the Netherlands, Saoudi Arabia, USA): 1. "Critical conversations: a user-centric approach to chatbots for history taking in the pediatric intensive care unit", by Candace Collins, James Fackler, Melissa Jerdonek Sacco and Maia Jacobs aims to explore the use of chatbots in pediatric medical settings, focusing on the need for emergent interventions in intensive care units. This works suggest a critical view on how to include users, such as patients, caregiver and clinicians to generate diagnostic reasoning and mitigate false information. Challenges of using chatbots in intensive medical care facilities are discussed and solutions for improvements are proposed. 2. "Experiment with ChatGPT: methodology of first simulation", by Oleg Shvets, Kristina Murtazin, Gunnar Piho and Martijn Meete indicate an experimental way to evaluate ChatGPT outputs in different assignment topics as compared with supervisor-provided feedback regarding answers' clarity, depth and relevance. Using students as participants, this work proves that both types of feedback were positively perceived, with a slight preference for the ChatGPT answers regarding their perceived ability to understand the topic of the given assignments. This contribution sheds light on the potential advantages of AI-generated feedback in educational settings, especially when contributing to topic clarification and when personalized feedback from supervisors is unavailable or limited. 3. "Exploring health professionals' views on the depiction of conversational agents as health professionals: A qualitative descriptive study", by A Luke MacNeill, Lillian MacNeill, Alison Luke and Shelley Doucet focuses on the role of conversational AI in health settings, showing the advantages and challenges of using conversational agents in the health care system in general. The results are valuable both for users and developers in offering further guidance and recommendations. 4. "Free word association analysis of students' perception of artificial intelligence", by Marvin Henrich, Sandra Formella-Zimmermann, Sebastian Schneider and Paul Wilhelm Dierkes is focused on students' perceptions of AI using semantic concept association. The authors explore how AI concepts are clustered and apparent in students' representations and what immediate applications they envisage for different assistance systems. The article critically examines the role of AI in the educational environment, the practical understanding of some very abstract terms at a student level, and offers reflections of how to facilitate AI literacy in schools and universities. 5. Intention to Use ChatGPT among Law Educators in Saudi Arabia", by Jawahitha Sarabdeen draws attention to another challenging topic, the use of common AI systems, such as ChatGPT, in professional domains, particularly ones with political implications, such as law education.The The authors explore the performance and effort expectations as well as facilitating conditions and behavioral intentions associated with adoption and use. The work draws attention to the importance of creating strong policies to regulate the proper acceptance and use of ChatGPT and similar AI applications by law educators and other key domains. The research also discusses the need for proper training for law educators, particularly in countries with a relatively recent history of technology use. This research topic offers key insights into the realm of user-chatbot interactions and their potential benefits for users, developers and policy makers. These relate to both opportunities and drawbacks. Based on the findings across all five studies presented here, we conclude that more longitudinal empirical studies in a variety of contexts and countries are needed. Chatbots, as humanlike text generators, might be considered as friends, foes, or both, and their pervasiveness in all spheres of social and professional life is made apparent in the works of the researchers included in this research topic.

Ähnliche Arbeiten