Short answer
Integrate structured opportunities for interdisciplinary teams to discuss, define, and evaluate AI fairness throughout the design and development lifecycle.
- Field
- User-Centred Design
- Source
- Academic Publication (2023)
- Method
- Qualitative research combining interviews and design workshops.
- Sample
- 23 industry practitioners from 17 companies.
- Evidence
- Moderate effect
Effective cross-functional collaboration is crucial for addressing fairness challenges in AI development by enabling practitioners to overcome differences in understanding, context, and evaluation. This user-centred design research insight is drawn from a 2023 study published in Academic Publication. Using Qualitative research combining interviews and design workshops. with 23 industry practitioners from 17 companies., researchers explored how this design variable affects real-world outcomes. The key design takeaway: Integrate structured opportunities for interdisciplinary teams to discuss, define, and evaluate AI fairness throughout the design and development lifecycle.
Bridging disciplinary gaps enhances AI fairness in design practice
Effective cross-functional collaboration is crucial for addressing fairness challenges in AI development by enabling practitioners to overcome differences in understanding, context, and evaluation.
Academic Publication · 2023
Key Findings
- 01Practitioners engage in 'bridging work' to reconcile differing perspectives on AI fairness across roles.
- 02When resources and incentives are lacking, practitioners leverage existing requirements (e.g., privacy) and evaluation norms (e.g., quantitative metrics) for fairness work, though this may be suboptimal.
- 03Significant 'invisible labor' is undertaken by practitioners to facilitate cross-functional collaboration for fairness.
Application
Design takeaway
Integrate structured opportunities for interdisciplinary teams to discuss, define, and evaluate AI fairness throughout the design and development lifecycle.
How to apply
In your design projects, actively seek out and facilitate discussions with individuals from different disciplines (e.g., engineering, ethics, marketing) to ensure a holistic approach to AI fairness.
Project actions
- 01When researching AI fairness, consider interviewing people with different roles in a company.
- 02Think about how different team members might understand 'fairness' differently and how to bridge those gaps.
Method & Evidence
Variables
Strengths & Limitations
Strengths
- +In-depth qualitative data from real-world practitioners.
- +Focus on practical challenges and opportunities.
Limitations
The study involved a limited number of companies and practitioners, so the findings might not apply to all industry contexts.
Reliability & validity
The study's validity is strengthened by the use of multiple data collection methods (interviews and workshops) and a diverse sample of practitioners. Reliability could be enhanced by further triangulation of findings across a larger, more varied set of organizations.
Think critically
How might organizational structures and incentives directly impact the ability of design teams to effectively collaborate on AI fairness?
Design Principles
"Foster collaborative intelligence by creating shared understanding and communication channels across diverse expertise."
As AI systems become more integrated into various aspects of life, ensuring their fairness is paramount. This research highlights that siloed expertise can hinder fairness efforts, emphasizing the need for design practices that foster interdisciplinary communication and shared understanding.
What This Means for Your Design
When building AI systems, people from different job types need to talk to each other to make sure the AI is fair. Sometimes, people have to do extra work to help everyone understand each other, especially if the company doesn't give them enough support.
How to use in your project
- 1.Reference this study when discussing the importance of interdisciplinary collaboration in your design process, particularly for complex ethical considerations like AI fairness.
Add to My Project
Quick Cite
Paragraph starter
This research highlights that effective cross-functional collaboration is a significant factor in addressing AI fairness. Practitioners often engage in 'bridging work' to overcome differences in understanding and evaluation across roles, especially in environments lacking dedicated resources or incentives for fairness. This underscores the need for design projects to proactively foster interdisciplinary communication and shared understanding to ensure equitable AI outcomes.
Source
Academic Publication
Investigating Practices and Opportunities for Cross-functional Collaboration around AI Fairness in Industry Practice
journal · 2023
View sourceQuestions About This Research
- What does the research say about bridging disciplinary gaps enhances ai fairness in design practice?
- Integrate structured opportunities for interdisciplinary teams to discuss, define, and evaluate AI fairness throughout the design and development lifecycle. Evidence: Academic Publication (2023).
- Why does "Bridging disciplinary gaps enhances AI fairness in design practice" matter for design?
- As AI systems become more integrated into various aspects of life, ensuring their fairness is paramount. This research highlights that siloed expertise can hinder fairness efforts, emphasizing the need for design practices that foster interdisciplinary communication and shared understanding.
- How can designers apply this research?
- Integrate structured opportunities for interdisciplinary teams to discuss, define, and evaluate AI fairness throughout the design and development lifecycle.
- What were the main findings?
- Practitioners engage in 'bridging work' to reconcile differing perspectives on AI fairness across roles.. When resources and incentives are lacking, practitioners leverage existing requirements (e.g., privacy) and evaluation norms (e.g., quantitative metrics) for fairness work, though this may be suboptimal.. Significant 'invisible labor' is undertaken by practitioners to facilitate cross-functional collaboration for fairness.
- What research method was used?
- Qualitative research combining interviews and design workshops. with 23 industry practitioners from 17 companies..
- How strong is the evidence?
- Evidence strength is rated Moderate effect, based on a 2023 journal from Academic Publication.
- What should I do differently in my next project?
- In your design projects, actively seek out and facilitate discussions with individuals from different disciplines (e.g., engineering, ethics, marketing) to ensure a holistic approach to AI fairness.
- What are the limitations?
- Findings may be specific to the companies and roles represented; the long-term effectiveness of 'piggybacking' tactics is uncertain.