Abstract
The paper presented by Smith et al. (2023) highlighted some interesting aspects regarding the use of ChatGPT in the teaching of social psychiatry. The authors have raised interesting points, particularly those related to the generalization of this educational possibility to other regions and languages. In Latin America, for example, the possibility of using ChatGPT could help create cases in Spanish or Portuguese, which could be used with medical students or residents. Nevertheless, there are some limitations to its use in this region, because it is necessary to train students in the evaluation of cases of patients who could use indigenous languages, such as Guarani in Paraguay, which is spoken by a significant part of the population (Aboaja et al., 2022). It has also been reported that even when searching for information online in this region, people more often use terms in indigenous languages than in Spanish (Singh et al., 2007). This situation has been replicated in other countries with significant indigenous populations and immigrants from different original languages. There has already been concern about this with more spoken languages such as French or Arabic, which are much more used online (Seghier, 2023). This situation could represent a new source of inequality in access to technology in societies that have already plagued structural differences in their opportunities.
One of the tasks given to AI in this study was to generate a clinical vignette about a case in social psychiatry. The ability to easily generate case vignettes in an educational framework is emphasized in this article. In recent years, the amount of work in educational institutions has increased substantially, making it necessary for academics to balance research, teaching, and healthcare provision in relatively compressed schedules (Bucklin et al., 2014). The availability of a teaching tool like ChatGPT that is able to generate case vignettes in a narrow space of time could mean for these professionals a lot of new time they could invest in their research projects and see patients.
In psychiatry, as the article stated, these case vignettes are particularly complex, with the need to include data about various aspects of the life of the patient in order to make an adequate formulation by the student. In particular, with the use of the biopsychosocial model, the vignette must provide some insight into the relational framework of the patient, their background, their workplace, the economic situation of the patients, and their family. The instructor can use this technology to generate a base clinical vignette and then propose more complex tasks to the student. For example, there are discussions about the ethical implications related to certain contexts, such as patients who suffer exploitation or sexual abuse. It could also be used to discuss critical incidents, such as problems related to the provision of adequate and up-to-date medical care in areas with reduced resources in the health systems, such as those encountered in South America. However, there have been concerns about how algorithms could reinforce economic or racial discrimination without appropriate control from medical professionals (Keskinbora, 2019).
Continuing with ethical considerations, there is a possibility of students using AI as a tool for plagiarism, particularly in tasks related to making essays or reports (Sallam, 2023). One alternative to students using this kind of tool is to increase the use of oral discussions or debates instead of writing documents in order to increase analysis and critical competency. The use of this kind of technology could paradoxically increase person-to-person interaction in order to limit the negative effects, instead of people learning only from machines or an AI, as some critics imply (King, 2023).
Regarding the other ways in which ChatGPT proposed to support social psychiatry teaching, it is very interesting to use it as a way to obtain resources and actualized information. There have been experiences with medical students and residents using mobile apps and programs in clinical settings (Fournier, 2022). Even with regard to psychiatry, previous studies have highlighted the perceived usefulness seen by students of these apps in learning (Lau & Kolli, 2017). The use of ChatGPT would be a step further in the integration of assistance in clinical decisions. Ultimately, it would still require a human to interpret the information, but this kind of app could dramatically improve the care provided, which may reduce the number of mistakes or enable more time in the training redirected to train future professionals on ethical dilemmas and interpersonal skills.
While there are many detractors for the use of ChatGPT, the advancement of technological development does not stop just because some people do not like it (Yadava, 2023). The reality is that ChatGPT and other AIs will continue to appear more frequently, with a greater capacity for analysis and synthesis, and will become increasingly widespread in different areas. Their application in the teaching of social psychiatry should be taken as a sign of new times and should be an opportunity for all those involved in the educational environment to use new technologies appropriately. Any new discovery has more promising aspects and other more obscure ones; however, if we are able to properly train the people who use these tools, we can put them to use that results in the benefit of patients and society.
