Interdisciplinary Collaboration is Key for Human-Centered AI in the Workplace
Category: User-Centred Design · Effect: Strong effect · Year: 2023
Developing and implementing effective human-centered AI in work environments requires a unified approach that integrates insights from diverse fields like Human Factors, Psychology, HCI, Information Science, and Adult Education.
Design Takeaway
To design truly human-centered AI for work, integrate perspectives from human factors, psychology, HCI, information science, and adult education to ensure systems support user capabilities, autonomy, and learning.
Why It Matters
As AI becomes more prevalent in professional settings, understanding its impact on users from multiple perspectives is crucial. This interdisciplinary view ensures that AI systems are not only technically sound but also ethically aligned, usable, and supportive of human capabilities and autonomy.
Key Finding
By examining how different fields approach human-centered AI, this research found that a combined perspective is necessary to effectively integrate AI into work, addressing user capabilities, autonomy, learning, and information needs.
Key Findings
- Disciplinary differences exist in the scope and conceptualization of HCAI.
- An interdisciplinary approach is essential for the successful development, transfer, and implementation of HCAI.
- Key aspects for successful HCAI include human capability and controllability (HFE), autonomy and trust (Psychology/HCI), learning and teaching designs (Adult Education), and information behavior/literacy (Information Science).
- A Synergistic Human-AI Symbiosis Theory (SHAST) is proposed as a foundational framework.
Research Evidence
Aim: How can an interdisciplinary theoretical framework guide the development and implementation of human-centered AI in the workplace?
Method: Comparative analysis and theoretical synthesis
Procedure: The research systematically mapped and compared conceptualizations of human-centered AI (HCAI) from Human Factors and Ergonomics (HFE), Psychology, Human-Computer Interaction (HCI), Information Science, and Adult Education. It analyzed their normative, theoretical, and methodological approaches to HCAI and synthesized these into a proposed interdisciplinary theory.
Context: Artificial Intelligence in the workplace
Design Principle
Human-centered AI design necessitates an interdisciplinary approach that prioritizes user capabilities, autonomy, trust, and continuous learning.
How to Apply
When designing AI-powered tools for professional environments, actively seek input from experts in ergonomics, psychology, and information science to ensure the AI enhances, rather than hinders, human performance and well-being.
Limitations
The proposed theory is foundational and requires further empirical validation across various work contexts. The specific methodologies and normative stances of each discipline may present challenges in full integration.
Student Guide (IB Design Technology)
Simple Explanation: To make AI work well for people in their jobs, we need experts from different fields (like engineers, psychologists, and educators) to work together. This helps ensure the AI is useful, safe, and easy to learn.
Why This Matters: Understanding how different fields view human-centered AI helps you design more comprehensive and effective solutions that consider the full spectrum of user needs and potential impacts.
Critical Thinking: To what extent can a single designer or a small team truly embody the necessary interdisciplinary expertise for human-centered AI, or is collaboration with external specialists always a prerequisite?
IA-Ready Paragraph: The development of effective human-centered AI in professional settings is critically dependent on an interdisciplinary approach, integrating insights from fields such as Human Factors, Psychology, Human-Computer Interaction, Information Science, and Adult Education. This holistic perspective ensures that AI systems address not only functional requirements but also crucial user aspects like capability, controllability, autonomy, trust, and learning, as highlighted by Mazarakis et al. (2023).
Project Tips
- When researching AI applications, consider the different perspectives of various disciplines.
- Justify the need for interdisciplinary collaboration in your design project's research phase.
How to Use in IA
- Reference this paper when discussing the importance of interdisciplinary approaches in your design project's research or justification sections.
- Use the identified key aspects (capability, autonomy, trust, learning) as criteria for evaluating your design's human-centeredness.
Examiner Tips
- Demonstrate an awareness of the multidisciplinary nature of human-centered design, especially when dealing with complex technologies like AI.
- Ensure your design rationale clearly articulates how user needs from various domains have been addressed.
Independent Variable: ["Disciplinary perspectives on HCAI","Integration of HFE, Psychology, HCI, Information Science, Adult Education"]
Dependent Variable: ["Effectiveness of HCAI development and implementation","User acceptance and performance with AI systems"]
Controlled Variables: ["Workplace context","Type of AI technology"]
Strengths
- Provides a comprehensive overview of different disciplinary views on HCAI.
- Proposes a novel theoretical framework (SHAST) for interdisciplinary AI research.
Critical Questions
- How can the proposed SHAST be practically applied and tested in real-world design projects?
- What are the potential conflicts or trade-offs when integrating normative stances from different disciplines?
Extended Essay Application
- Investigate the application of a specific aspect of HCAI (e.g., user autonomy) in a chosen workplace context, drawing on relevant disciplinary theories.
- Propose an interdisciplinary framework for evaluating the human-centeredness of an existing AI system in a professional setting.
Source
What is critical for human-centered AI at work? – Toward an interdisciplinary theory · Frontiers in Artificial Intelligence · 2023 · 10.3389/frai.2023.1257057