Rapid Mobile App Development for Generative AI Agents on MIT App Inventor
Project Overview
The document explores the integration of generative AI in education, particularly through the MIT App Inventor platform, which facilitates the creation of AI-driven mobile applications. It highlights three key applications aimed at sustainability, productivity, and safety, showcasing how generative AI can improve educational tools and promote community engagement. The discussion includes the challenges encountered during the integration of AI functionalities, emphasizing the necessity of user-centered design in the development process to ensure these applications effectively meet the needs of educators and learners. Overall, the findings emphasize the potential of generative AI to transform educational experiences, enhance learning outcomes, and address pressing societal issues through innovative technological solutions.
Key Applications
AI Support Tools
Context: Educational applications aimed at enhancing user engagement and improving processes, including study habits for secondary students, reporting local environmental issues, and providing community safety resources. These applications target users in various educational and community contexts, promoting collaboration, sustainability awareness, and emergency assistance.
Implementation: Developed using MIT App Inventor, these applications leverage AI tools like ChatGPT for features such as generating practice questions, providing tailored study tips, identifying environmental threats, and connecting users with local emergency resources. The methodology focuses on user engagement and personalized feedback to adapt functionalities to user needs.
Outcomes: Enhanced study habits and information retention through personalized techniques, increased environmental awareness and community engagement, and improved safety through centralized resource access. The applications foster collaboration and motivation among users, promoting proactive participation.
Challenges: Common challenges include limitations on AI feature integration (e.g., quota restrictions on question generation), dependence on user engagement for effectiveness, and concerns regarding user data privacy and ethical implications in handling sensitive information.
Implementation Barriers
Technical Limitations
MIT App Inventor has restrictions on the number of API calls, which can limit the functionality of generative AI features.
Proposed Solutions: Developers are encouraged to reload the app or optimize the flow to manage API usage effectively.
User Engagement
The effectiveness of educational apps relies heavily on user input and engagement, which can be inconsistent.
Proposed Solutions: Incorporating user feedback mechanisms and incentives to encourage active participation and improve app functionality.
Data Privacy Concerns
Safety applications must manage user data responsibly to maintain trust and ethical standards. Clear communication about data usage and privacy policies is essential to maintain user trust.
Proposed Solutions: Implement clear communication about data usage and privacy policies to users.
Project Team
Jaida Gao
Researcher
Calab Su
Researcher
Etai Miller
Researcher
Kevin Lu
Researcher
Yu Meng
Researcher
Contact Information
For information about the paper, please contact the authors.
Authors: Jaida Gao, Calab Su, Etai Miller, Kevin Lu, Yu Meng
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