Researching Translation Topics: AI for Locating Sources

Conceptión Godev

Locating References, Perplexity, Copilot, Gemini

“I don’t fear it… yet. I’m cautiously curious to learn more and understand how my courses can integrate AI in an ethical way so that my students may see connections between both how AI can be used for their academic goals and how they may end up using AI in their future professional environment.”

Conceptión Godev

Recently, my undergraduate students and I have used NaturalReader, which is an AI-powered text-to-speech application, very useful for language learners to study pronunciation and enunciation. However, generative AI in its latest incarnations as Large Language Models such as ChatGPT is a different type of AI in its potentially far-reaching impact at all levels of activities that rely on producing text. As many of my colleagues at UNC Charlotte and in academia in general, I’m learning about AI, its capabilities and shortcomings. I don’t fear it … yet. I’m cautiously curious to learn more and understand how my courses can integrate AI in an ethical way so that my students may see connections between both how AI can be used for their academic goals and how they may end up using AI in their future professional environment. 

Because it is not possible for me to make a wholesale implementation of a tool that I do not completely understand at the moment, I chose to integrate an experimental task into the final research project of my course TRAN 6602: Linguistics for Translators. The report on the experimental task will become an appendix to the final project. The task will be completed only by students who are willing to do it. At the moment, three students will be completing the task. The goal of the task is to explore to what extent three AI chatbots, namely Perplexity, Copilot, and Gemini, can complement the research process of checking databases in the library to locate references on a specific topic.

 Here are the instructions that the students received:

1. Use these platforms (Perplexity, Copilot, and Gemini) to locate authoritative references (peer-review, dissertations, theses, no textbooks or encyclopedias) on one of the subtopics of your final project in TRAN 6602: Linguistics for Translators.
2. Ask one or two questions with the objective of receiving a list of references. For example, here is a possible prompt: Give me a list of five peer-reviewed references on how translators use AI.
3. Ask exactly the same question or questions of the three systems you will use.
4. Report on:
   1. your findings on how the AI systems helped or did not help, and 
   2. how that interaction was, better, worse or complemented your interaction with the library databases in terms of quantity of resources, the quality of sources and the time on task. You may use a table to show the prompts and the information returned by each system.
5. Include your report in an appendix to your final project. The appendix needs to be placed after the list of references.

Since the final project will not be complete until the end of Spring 2025, the students’ reports are not yet available. This use-case, however, is easy to replicate and provides a conservative way to observe AI capabilities for the purpose of a very specific research task that usually is part of the process of writing research papers.