Abstract
As artificial intelligence (AI) continues to increase its presence and accessibility within education, the need to address AI's impact on assignment design and the production of original coursework is heightened. Within the context of an undergraduate language and culture course, this chapter thus offers reflections on the integration of AI tools and their effect on shaping assessment methods. The authors also highlight that there indeed remains a great need for continued research in the realm of AI and education going forward, especially where enhanced AI-detection technologies, institutional policies, academic rigour, and learner expressiveness are concerned.
TopCurrent Debate Surrounding Ai
Current hesitations around the use of AI in education are simply the next phase of a decades-long discussion on the feasibility and necessity of bringing new technologies into the classroom which, in recent years, has also engaged topics such as text messaging (Carrington, 2005) and machine translation (Urlaub & Dessein, 2022).
With regards to the language classroom, Urlaub and Dessein (2022) diffuse the perceived disruption caused by machine translation, such as Google Translate, to the acquisition of language-learning outcomes by drawing an intriguing comparison with pocket calculators in Mathematics classrooms, which became a widespread practice in schools only once the S.A.T. permitted their use in 1994. Initial debates in the 1970s hinged upon whether the availability of pocket calculators in the mathematics classroom would lead to a loss of basic arithmetic skills among students. Urlaub and Dessein (2022) point to a shift in instructor perception of the pocket calculator, as it went from being considered an impediment to a learning tool that could optimize learning outcomes. Analogously, they argue that a targeted and thoughtful approach toward machine translation in the language classroom could enhance learning outcomes. The prevailing concern is that the frequent use of machine translation could prevent students from accumulating the skills necessary to produce written communication in the target language without the support of aids (51). Urlaub and Dessein (2022) admit that there is a real danger that tools such as Google Translate can contribute to a reductionistic perception of language and language learning. They underscore that if language proficiency is “treated as a tool, it reduces human beings to speakers exchanging messages in crude manners that are agnostic of the sociocultural embeddedness of message and speaker” (p. 57). Naturally, this concern has implications for oral and written communication across academic disciplines. However, the possibility that communication in a foreign language could be reduced to mere “translation” particularly stands out in the language classroom. In order to combat this outcome, Urlaub and Dessein (2022) argue that we must understand language proficiency as a nuanced and context-sensitive form of communication that reflects and represents the individual within society. Moreover, machine translations’ inability to harness cultural nuance (through, for example, idiomatic expressions)1 in fact limits the depths to which it is beneficial in achieving language competence.
Key Terms in this Chapter
Generative AI: Artificial Intelligence technologies that generate responses based on inputted prompts.
Multimodal: Coursework or activities which are enhanced through a variety of formats (e.g. audio-visual assignments, in-person and virtual delivery of lecture content).
Cultural Competence: Increasing individuals’ cross-cultural awareness and appreciation. Achieving cultural competence is seen as inextricably linked to language proficiency.
Digital Literacy: Honing the skills needed to effectively navigate today’s technology-drive world.
Reflective Assignments: We define reflective assignments as a blend of students’ considerations on the ways in which their in-class and lived experiences intersect with their academic research and analyses of scholarly sources.
Italian Cultural Studies: Examining various aspects of Italian culture (art, fashion, cinema, cuisine, etc.) within historical context.
LLMs: Large Language Models (LLMs) are sophisticated AI systems trained on vast amounts of textual data, allowing them to learn complex linguistic patterns. The versatility of LLMS makes them useful tools that could be applied to language processing tasks such as translation, answering questions, and providing summaries. Within the umbrella of Generative AI tools available to the public, ChatGPT, developed by OpenAI, falls under the category of an LLM-based chatbot. Other LLM-based chatbots include Bard, developed by Google, and Ernie, developed by Baidu. While Google Translate is an LLM, it is not an LLM-based chatbot.
Experiential Learning: Hands-on learning and reflection through direct contact with the subject matter.