Daily Note: Explainable Artificial Intelligence and Tangible Embodied Interaction

These notes are a summary of concepts presented in “From ”Explainable AI” to ”Graspable AI.”

Maliheh Ghajargar, Jeffrey Bardzell, Alison Smith Renner, Peter Gall Krogh, Kristina Höök, David Cuartielles, Laurens Boer, and Mikael Wiberg. 2021. From ”Explainable AI” to ”Graspable AI”. In Proceedings of the Fifteenth International Conference on Tangible, Embedded, and Embodied Interaction (TEI ’21). Association for Computing Machinery, New York, NY, USA, Article 69, 1–4. https://doi.org/10.1145/3430524.3442704

  1. Explainable AI
    • Explores mechanisms to explain or expose AI’s inner workings to increase understanding and trust
    • Types of explanations
      • Global explanations – explain models as a whole
      • Local explanations – rationalize individual predictions
    • User benefits
      • Enhances understanding of AI systems
      • Enables users to improve or influence AI by recognizing errors
    • Interactive Machine Learning (Human-in-the-loop)
      • Supports iterative feedback to refine models
  2. Tangible Embodied Interaction
    • Tangible User Interfaces
      • Merge computing systems with everyday physical objects and spaces
    • Evolution of tangible embodied interaction
      • Initially focused on control and manipulation
      • Expanded to include representation and physicalization of data
    • Embodied interaction
      • Examines how computing alters our perception of physical reality
  3. Material-Centered Interaction Design
    • Emphasis on material aspects of interaction
    • Soma design
      • A body-centric approach incorporating movement and holistic interaction
  4. Forms & Form-Giving in Interaction Design
    • Forms convey meaning through perceived affordances
    • Design process considerations
      • Forms result from synthesizing context, user needs, and materials
    • Graspable forms
      • Unified wholes conveying meaning, message, or state
      • Often self-explanatory and intuitively understandable
      • Change over time
  5. Graspable AI
    • Opportunities for democratic artificial intelligence and machine learning engagement
      • Encourages criticism, participation, and mutual understanding
    • Graspable interaction criticism
      • Examines relationships between physical and intellectual dimensions of AI
    • Key themes in Graspable AI
      • Tangible manipulation
      • Spatial interaction
      • Embodied facilitation
      • Expressive representation