17–19 May 2024
Meijo University Nagoya Dome Campus
Asia/Tokyo timezone

Beyond Basics: A Five-Week Deep-Dive into AI for University Language Students

18 May 2024, 10:10
30m
DN 412 (North Building)

DN 412 (North Building)

Practice-based Presentation (30 minutes) AI for Teaching DN 412: AI for Teaching

Speakers

Mr Euan Bonner (Kanda University of International Studies) Erin Frazier (Meiji University) Ryan Lege (Kanda University of International Studies)

Description

How do we handle Large Language Model (LLM) AI in the classroom? Many approach it by creating policies that either restrict, ban, or limit its use to specific AI-focused activities. This can be effective in the short term, however, many underestimate the effect this technology will have on language learners’ futures as global digital citizens. Additionally, if AI is approached through limited implementations, like single activities, this can lead to an inflated sense of what the technology can do while underselling its limitations. There is a need to dedicate extended class time to covering AI so that students understand the real affordances and pitfalls of the technology to make informed decisions about when and how to use it. Accordingly, this presentation will discuss the creation and implementation of a 10-lesson unit that has since been integrated into 1st and 2nd-year courses at multiple universities.

The unit introduces students to the AI landscape, starting with an overview of AI's history and explaining how LLMs work, highlighting their differences from traditional AI. It teaches students prompt engineering, then tasks them with creating a range of AI-generated media, including art, videos, audio, HTML pages, and web games. The lessons are structured following Highland’s (2009) teaching-learning cycle. The unit features a final project where students design a local newspaper using AI. The evaluation focuses on their reflection and the effectiveness of their prompts rather than the product itself. Additionally, the unit delves into ethical considerations like fake news, job displacement, and academic integrity.

Additionally, the researchers developed a bespoke AI teaching assistant web-based chat application that was trained on the lesson plans and materials. The application allows students to converse with the AI assistant to review materials, get help with lesson vocabulary, and ask questions both in English and their native languages.

Keywords AI, Large Language Models (LLMs), Curriculum Design, AI Tutors

Primary authors

Mr Euan Bonner (Kanda University of International Studies) Erin Frazier (Meiji University) Ryan Lege (Kanda University of International Studies)

Presentation materials

There are no materials yet.