AERC Hosts: Alina von Davier "Human-Centered AI and Trust Building in Language Learning and Testing Systems"

Abstract:

Artificial intelligence (AI) has revolutionized the landscape of language learning and testing in recent years. The rapid development of AI technologies has led to the creation of advanced language learning platforms and assessment tools that can adapt to individual learners' needs, providing personalized tests and feedback. As AI systems become increasingly integrated into educational contexts, it is essential to explore the concept of human-centered AI and the role of trust in these systems.  Human-centered AI refers to the design and implementation of AI technologies that prioritize human values, needs, and goals. In the context of language learning and testing, this means creating AI systems that align with pedagogical principles, ethical considerations of educators and learners, and test takers' experience and access. Trust in AI systems is crucial for the widespread adoption of these technologies, as users must feel confident in the reliability, fairness, and accuracy of the AI-driven recommendations and assessments.  

In this presentation, I will discuss the current state of human-centered AI in language learning and testing, addressing the design principles and ethical considerations that must be taken into account when developing AI-driven tools. I will also discuss the features of human-in-the-loop AI as one approach to support human-centered AI processes and decisions in learning and assessment systems. Furthermore, I will explore the notion of trust in AI systems, examining factors that influence user trust, such as transparency, explainability, and accountability.  By emphasizing the importance of human-centered AI and trust in AI systems, we aim to encourage the future development of more effective, inclusive, and ethically sound language learning and testing technologies. I will illustrate these principles with the Duolingo English Test Responsible AI Standards Burstein, 2023) and their role in supporting both the test development process and its audit.

RSVP: AERC-info@unimelb.edu.au