How Experienced Designers of Enterprise Applications Engage AI as a Design Material

Document Title

How Experienced Designers of Enterprise Applications Engage AI as a Design Material

Document Information

• Subject Area: Human-Computer Interaction (HCI), Artificial Intelligence (AI), and User Experience (UX) Design
• Keywords: User Experience Design, Artificial Intelligence, Machine Learning, Human-Computer Interaction, Service Design, System Design, Design Tools, Data Visualization, Design Blueprint, Interdisciplinary Collaboration

Research Background and Questions

• Research Background: As AI's importance in product and service user experience continues to grow, HCI research has begun to view AI as a design material—a technical capability that designers can leverage to envision new opportunities. However, in practice, designers often face difficulties in understanding AI's capabilities and limitations, which hinders their ability to innovate.
• Research Questions:
  1. How do designers identify and design new AI-based opportunities and interactions?
  2. What challenges must designers in interdisciplinary AI teams overcome to drive enterprise-level AI innovation?
  3. How can collaboration between designers and data scientists be facilitated?
• Research Motivation and Related Work:
  • Current research indicates that adaptive user interfaces (AUI) in enterprise applications are a potential area where designers can identify opportunities to enhance UX with AI, but this is often overlooked in practice.
  • Collaboration between designers and data scientists is hindered by a lack of shared language and workflows. Boundary objects that bridge these two fields may facilitate interdisciplinary communication.

Solutions

• Methods and Design:

  1. Through design workshop sessions, explore how experienced designers in AI teams uncover design opportunities with AI and address design challenges in team collaboration.
  2. The workshops involve multiple design and technical roles both within and outside enterprises, discussing key issues in the design process across three stages (early discovery, mid-stage definition and development, and late-stage delivery).
  3. Use tools such as service blueprints, data visualizations, and logic flow diagrams to understand and communicate the relationship between design and AI systems.

• Innovations:

  1. Expands the concept of "AI as a design material," focusing on how designers innovate at the service and system levels rather than just the UI level.
  2. Highlights the importance of collaboration between designers and data scientists, proposing new design tools and methods (e.g., service blueprints with added data lanes).

Research Outcomes

• Specific Findings:

  1. Scope of Designers' Influence: Designers' innovations are not limited to interface-level task optimization but also include enhancements at the service process and system goal levels.
  2. Main Barriers to AI Innovation: Designers do not face difficulties in identifying AI-enhanced opportunities, but proving business value and return on investment becomes the primary obstacle.
  3. Interdisciplinary Collaboration and Boundary Objects: Tools such as data visualizations, blueprints, and logic diagrams serve as boundary objects in collaboration, helping bridge the gap between design and data science fields.

• Comparison with Existing Solutions:

  • Current discussions mainly focus on designers' roles at the UI level, whereas this study demonstrates designers' broader potential in system and service design.
  • Provides a new perspective on how AI design in enterprise applications differs from consumer application design.

• Experimental or Evaluation Results:

  1. Use of Tools and Methods: Designers employed enhanced service blueprints, AI creativity matrices, and data-driven service design canvases, effectively facilitating team communication and alignment.
  2. Case Analysis: The study summarized how designers achieved a deep understanding of data and AI systems through collaboration and identified actionable AI innovations.

• Limitations and Future Directions:

  1. Limitations: The study sample includes only experienced designers, lacking broader coverage of other enterprise contexts and roles; privacy and ethical issues were not deeply explored.
  2. Future Directions:
     • Develop tools to support AI cost and value estimation.
     • Explore the expanded role of designers in AI teams, particularly in problem framing and systems thinking.
     • Design and evaluate boundary objects that promote multidisciplinary collaboration, supporting areas such as fairness, transparency, and privacy.

Conclusion

This study offers a unique perspective on how designers in enterprise-level interdisciplinary teams utilize AI as a design material. By identifying designers' roles in AI-driven innovation and the main obstacles they face, the research provides new directions for improving design tools, methods, and collaborative practices, while also laying the groundwork for future studies.