Generative Co-Learners: Enhancing Cognitive and Social Presence of Students in Asynchronous Learning with Generative AI

TL;DR

This work introduces Generative Co-Learners (GCL), a multimodal AI-assisted system that uses generative agents as co-learners to augment cognitive and social presence in asynchronous learning. Through a 12-participant study with programming tutorials, GCL demonstrated improvements in cognitive engagement and social interaction, including richer notes, multimodal feedback, and believable agent behaviors. While learning gains (pre- to post-quizzes) improved with GCL, they did not significantly surpass a baseline system in short sessions, highlighting the need to explore active learning and longer studies. The paper discusses audience effects, vicarious learning, AI-generated feedback, and ethical considerations, proposing future work to scale customization, ensure accuracy, and address cognitive load in AI-enabled asynchronous education.

Abstract

Cognitive presence and social presence are crucial for a comprehensive learning experience. Despite the flexibility of asynchronous learning environments to accommodate individual schedules, the inherent constraints of asynchronous environments make augmenting cognitive and social presence particularly challenging. Students often face challenges such as a lack of timely feedback and support, a lack of non-verbal cues in communication, and a sense of isolation. To address this challenge, this paper introduces Generative Co-Learners, a system designed to leverage generative AI-powered agents, simulating co-learners supporting multimodal interactions, to improve cognitive and social presence in asynchronous learning environments. We conducted a study involving 12 student participants who used our system to engage with online programming tutorials to assess the system's effectiveness. The results show that by implementing features to support textual and visual communication and simulate an interactive learning environment with generative agents, our system enhances the cognitive and social presence in the asynchronous learning environment. These results suggest the potential to use generative AI to support students learning at scale and transform asynchronous learning into a more inclusive, engaging, and efficacious educational approach.

Figures (7)

• Figure 1: An overview of GCL user interface. There are four panels: (A) a main video panel for displaying the learning content; (B) a function panel with a text and code editor; (C) a co-learners panel displaying the video and screen share of simulated co-learners in asynchronous learning environments powered by generative AI; and (D) a chat panel for users to communicate with co-learners.
• Figure 2: A brush feature enabling users to highlight a specific area in the video and pose related questions to the co-learners.
• Figure 3: A text-based chat function that enables users to (a) send text messages to co-learners and ask questions related to the learning material. (b) When the co-learner responds to the user's message, the generative agents select the appropriate active action and display it on an enlarged responsive screen. (c) The text response will be displayed in a private chat window
• Figure 4: An audio chat function that enables users to (a) send voice chat messages to co-learners by clicking on the audio chat button. (b) When the co-learner responds to the user's message via audio, the generative agents select the appropriate active action and display it on an enlarged responsive screen. (c) The text from the audio response will be displayed in a private chat window
• Figure 5: User can view the co-learner generated (a) notes, (b) profile, and select preferred learner's tone, interaction style, or characteristic in (c) customization menu