Creative Wand: A System to Study Effects of Communications in Co-Creative Settings
Project Overview
The document explores the C REATIVE-WAND framework, a customizable system aimed at fostering co-creation between AI and human users in educational and creative contexts. It highlights the significance of effective communication in mixed-initiative systems, where both human designers and AI agents collaboratively influence creative outputs. A study focusing on storytelling demonstrates how the framework enables users to articulate their creative intentions, enhancing interaction with AI and leading to improved creative results. However, it also uncovers challenges such as user frustration and difficulties in achieving specific goals. Overall, the findings suggest that while generative AI has the potential to enrich educational experiences through enhanced creativity and collaboration, addressing user challenges is crucial for optimizing these interactions.
Key Applications
C REATIVE-WAND Framework
Context: Used for co-creative storytelling in educational settings, targeted at students or users interested in creative writing and AI interaction.
Implementation: The framework allows for plug-and-play integration of generative models and communication channels within a chat-based interface, facilitating interaction between users and AI.
Outcomes: Improved goal completion rates in storytelling tasks and enhanced user satisfaction through effective communication strategies.
Challenges: User frustration due to limited communication options and difficulty in achieving specific sub-goals.
Implementation Barriers
Usability Issues
Users may feel frustrated due to the cognitive load and limitations in how they can interact with the AI system, leading to underreporting of goal completion. Limited types of communication may hinder users' ability to achieve their creative goals effectively.
Proposed Solutions: Future work should focus on adaptive, personalized co-creativity systems that can better handle user feedback and reduce cognitive load, incorporating a mix of communication strategies in co-creative systems to accommodate various creative needs.
Project Team
Zhiyu Lin
Researcher
Rohan Agarwal
Researcher
Mark Riedl
Researcher
Contact Information
For information about the paper, please contact the authors.
Authors: Zhiyu Lin, Rohan Agarwal, Mark Riedl
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