Skip to main content Skip to navigation

Synergizing Human-AI Agency: A Guide of 23 Heuristics for Service Co-Creation with LLM-Based Agents

Project Overview

The document explores the integration of generative AI, particularly Large Language Models (LLMs), in education and public services, with a focus on libraries. It highlights the collaborative nature of human-AI interactions and the challenges faced by non-experts in effectively utilizing AI technologies. A key component is the introduction of CoAGent, a tool aimed at enhancing service co-creation with LLMs, which underscores the importance of clearly defining roles and responsibilities between humans and AI. The findings reveal 23 heuristics that facilitate effective collaboration and emphasize the recognition of AI as an active participant in service delivery. Overall, the document underscores the transformative potential of generative AI in educational contexts, advocating for ethical guidelines and best practices to maximize its benefits while addressing the challenges associated with its implementation. Through these insights, it points to the necessity of fostering a cooperative environment where both educators and AI can work together to enhance learning experiences, ultimately leading to improved educational outcomes.

Key Applications

CoAGent - a three-module design tool for service co-creation with LLM-based agents.

Context: Public libraries, serving as community hubs for education and digital literacy.

Implementation: A participatory design process involving domain experts from public libraries to co-create and refine the tool.

Outcomes: Enhanced understanding of AI's capabilities, improved service delivery, and the establishment of 23 heuristics for collaboration with AI.

Challenges: Knowledge disparities between AI and non-AI experts, issues of control over AI outputs, and the need for continuous updates to AI systems.

Implementation Barriers

Knowledge Gap

Non-AI experts face challenges in understanding and utilizing AI tools due to limited technical expertise.

Proposed Solutions: Incorporate training programs and resources to enhance understanding, and use participatory design processes to involve users in AI development.

Ethical Concerns

Lack of ethical guidelines for AI deployment in public services may lead to mistrust and quality issues.

Proposed Solutions: Develop clear ethical design principles and guidelines to ensure responsible AI integration in public services.

Project Team

Qingxiao Zheng

Researcher

Zhongwei Xu

Researcher

Abhinav Choudhry

Researcher

Yuting Chen

Researcher

Yongming Li

Researcher

Yun Huang

Researcher

Contact Information

For information about the paper, please contact the authors.

Authors: Qingxiao Zheng, Zhongwei Xu, Abhinav Choudhry, Yuting Chen, Yongming Li, Yun Huang

Source Publication: View Original PaperLink opens in a new window

Project Contact: Dr. Jianhua Yang

LLM Model Version: gpt-4o-mini-2024-07-18

Analysis Provider: Openai

Let us know you agree to cookies