Overcome the Algorithm: A Generative AI Guide for Celebrating Diversity

957 

Session 

Non-language specific 

English 

Stephanie Knight  
Center for Applied Second Language Studies

Stephanie Knight  
Center for Applied Second Language Studies

Generative artificial intelligence (AI) has transformed the instructional tool landscape for world language teachers and learners. It provides a new space for learners to engage in real-time practice and provides educators new ways to approach feedback and engineering classroom activities. However, AI also merits critique for its capacity to reinforce standard language norms and dominant cultural frames. This dual reality has sparked rich debates in the language community about creation, communication, and the quality of interactions learners have within the tools. This session honors this complex reality by providing a ready-made guide for iterating AI prompts that generate culturally and linguistically diverse output. Based on the Intercultural, Pragmatic, and Interactional Competence framework, this guide will also be used to introduce open-access pragmatics repositories as a resource for editing AI-generated output to further enhance said cultural and linguistic diversity.

Participants will be able to evaluate the output created by artificial intelligence tools through a lens of cultural and linguistic diversity by using the Intercultural, Pragmatic, and Interactional Competence framework.

Participants will be able to utilize open-access pragmatics repositories to evaluate and manipulate the output generated by artificial intelligence tools.

Participants will be able to describe how to engineer and edit prompts to produce output that reflects the cultural and linguistic diversity of target language speakers.

Presenters will display an AI-generated conversation in both standard English and another English dialect. Participants will evaluate these conversations using a thinking routine in which they describe what they see, explain their opinions about it, and state any inquiries they have. During this discussion, the presenters will review literature critiquing the cultural and linguistic diversity inherent in typical AI output.

Next, presenters will share a guide for AI prompt generation based on the Intercultural, Pragmatic, and Interactional Competence framework. This guide supports engineering prompts that produce output that accurately reflects grammar, vocabulary, and communication strategies (e.g., saying goodbye appropriately given the context) in a target language. Presenters will also use the guide to introduce open-access pragmatics repositories as a resource for editing AI output to further ensure accurate representation of cultural and linguistic diversity in the classroom.

Participants will consider how the algorithms utilized in artificial intelligence (AI) technologies are likely to reinforce standard language norms and dominant cultural frames. A guide to iterate more culturally and linguistically diverse AI interactions based on the Intercultural, Pragmatic, and Interactional Competence framework will be shared.

All Audiences

Technology and Digital Learning