Semi-structured interviews were conducted in either Spanish or English via video conferencing software with 11 participants, including three program coordinators from McMaster University, Maastricht University, and Universidad del Rosario, one tutor from McMaster University, two students from Universidad del Rosario and five students from McMaster University.
4.1. The Promises and Pitfalls of Machine Translation
Recent innovations in artificial intelligence (AI), and specifically machine translation, show promise for the creation of more inclusive and equitable online learning spaces in global health by allowing learners to speak their native language. At its core, AI refers to the ability of machines to carry out complex tasks typically requiring human intelligence . Supported by AI, machine translation requires only a computer, smartphone, or device connected to the internet and allows for more intelligent and nuanced translations . While little is known about the mechanics and outcomes of courses using machine translation, Australian and Indonesian medical students engaging in online communication tools and platforms identified real-time machine translation as a tool that facilitated fluid conversations between non-English-speaking and English-speaking groups .
Preliminary findings illustrate the potential to capitalize on the strengths of machine translation to reduce inequities produced by teaching global health predominantly in English. Partnering students from McMaster University (Canada) and Universidad del Rosario (Colombia) together in virtual learning pods with tutors who are fluent in Spanish, groups communicated using Microsoft Translator: Conversations. McMaster students, often with limited skills in written and oral Spanish, communicated with Spanish-speaking group members using machine translation, and with the assistance of the tutor. Students were resourceful and adaptable in navigating multiple communication technologies, relying on WhatsApp for group communication, Google Docs to collaborate on assignments, Zoom or Google Meets for group meetings, Google Translate for text translation, and Microsoft Translator: Communications for live translation during group meetings. Students emphasized that the multilingual component of group work contributed to “improved communication skills and ability to work in groups” and described multilingual group work as “richer” than previous experiences with group work. Students also described learning about other cultures as an important outcome of multilingual group work, which was noted during interviews as “essential” for work in global and public health. Canadian students emphasized gaining greater understandings of health challenges in Colombia, while Colombian students noted the ability to work with global North institutions as crucial for future work with multilateral funding agencies.
Significant limitations with currently available machine translation technologies were noted across the interviews, with Microsoft Translator: Communications being described as “ineffective”, “time-consuming”, and “inaccurate”. In line with the literature on the digital divide , some students noted their age or lack of familiarity with technologies as key barriers in adopting machine translation software. Given these challenges, some groups abandoned translation technologies and relied solely on bilingual tutors and group members to translate. Students also underscored the challenges in translating cultural contexts that do not necessarily have a linguistic equivalent in English, resulting in local meanings being lost in machine translation. For example, Colombian students described difficulties in translating concepts from Indigenous traditional medicines, and often relied on terminology from Western medicine when translating to English.