Trust and ethical considerations in a multi-modal, explainable AI-driven chatbot tutoring system: The case of collaboratively solving Rubik's Cube

Kausik Lakkaraju,Vedant Khandelwal,Biplav Srivastava,Forest Agostinelli,Hengtao Tang,Prathamjeet Singh,Dezhi Wu,Matt Irvin,Ashish Kundu

2024-08-27

Abstract:Artificial intelligence (AI) has the potential to transform education with its power of uncovering insights from massive data about student learning patterns. However, ethical and trustworthy concerns of AI have been raised but are unsolved. Prominent ethical issues in high school AI education include data privacy, information leakage, abusive language, and fairness. This paper describes technological components that were built to address ethical and trustworthy concerns in a multi-modal collaborative platform (called ALLURE chatbot) for high school students to collaborate with AI to solve the Rubik's cube. In data privacy, we want to ensure that the informed consent of children, parents, and teachers, is at the center of any data that is managed. Since children are involved, language, whether textual, audio, or visual, is acceptable both from users and AI and the system can steer interaction away from dangerous situations. In information management, we also want to ensure that the system, while learning to improve over time, does not leak information about users from one group to another.

Computers and Society,Artificial Intelligence

What problem does this paper attempt to address?

The paper aims to address the following issues:

1. **Trust Issues**: On an AI-based multimodal collaboration platform (named ALLURE chatbot), ensuring a safe environment for children when collaborating with AI to solve Rubik's Cube problems, and making the "black box" communication process in AI algorithms transparent. Specifically, ensuring that all languages (text, audio, or visual) are acceptable and that the system can guide interactions away from dangerous situations.

2. **Information Management Issues**: Ensuring that the system, while continuously learning and improving over time, does not leak information from one user group to another, to protect privacy and security.

The paper constructs a multimodal interactive chatbot system (ALLURE) to help high school students collaborate with AI to solve Rubik's Cube problems, while addressing the two main research challenges mentioned above. Additionally, the paper explores how to use deep reinforcement learning and inductive logic programming to find Rubik's Cube solutions that are easy for high school students to understand and to communicate these explanations to users in a personalized manner.