Learning Theories Could Transform Human-Centered XAI
A new position paper on arXiv argues that integrating learning theories into Explainable AI (XAI) can enhance human agency and mitigate risks. The authors propose a learner-centered approach to design, assess, and evaluate AI explanations, addressing the growing complexity of large AI systems. The paper explores opportunities and challenges of this approach, aiming to evolve human-centered XAI practices.
Key facts
- Paper titled 'Using Learning Theories to Evolve Human-Centered XAI: Future Perspectives and Challenges'
- Published on arXiv under Computer Science > Artificial Intelligence
- Discusses infusing learning theories into the XAI lifecycle
- Argues learner-centered XAI can enhance human agency and ease risk mitigation
- Addresses challenges of explaining large, complex AI systems
- Focuses on why and what to explain in AI transparency
- Builds on past work in human-centered XAI
- Submission history and references included on arXiv
Entities
Institutions
- arXiv