PROJECT SUMMARY
Critical Capabilities for Inclusive AI
Focus Areas: News & Media, Mobilities, Social Services, Health
Status: Active
Inclusive AI is related to, but distinct from Responsible AI and the ethical principles and governance frameworks currently in development. At base it involves ensuring that all members of society benefit from AI tools and ADM systems and can participate in their design or respond to their deployment. We see capabilities – machine and human – as central to how inclusive AI might be achieved.
While much of the research focus is currently targeting the features, functions and ‘use cases’ of LLMs and other AI model types, not enough emphasis is placed on the ‘human factors’ or the co-learning and socialisation taking place in real-world settings and among different groups using these tools and systems.
This project addresses the knowledge, skills and literacies – the critical capabilities – needed to achieve inclusive AI in Australia. It will work with research partners, consumers and communities to better understand the capabilities and resources needed to access and use AI tools including Generative AI. Central to the project is the AI Capabilities Lab, a platform and space to experiment, observe and evaluate the use of new AI tools with our industry partners and members of the public.
Through the AI capabilities Lab and participatory research methods, the project will build an evidence base about the shifting lines of expertise, knowledge and decision making in organisational and everyday life settings as people begin to use AI tools.
PROJECT OBJECTIVES
- Develop a model of AI capability and literacy for AI inclusion, tested with key domain areas and target populations;
- Generate empirical evidence about the way people and organisations are using AI tools, and their potential for alleviating or deepening digital inequalities;
- Co-design resources with partner organisations and their communities and consumers to enhance inclusive AI literacy and capability and foster responsible forms of ‘social governance’ for AI use and ADM processes; and
- Develop and test AI capability and usage metrics, test and evaluate AI tools and systems through user studies and explore meta-evaluation approaches for targeted AI tools and applications.
MORE INFORMATION
PUBLICATIONS
Relational Ethics in Health Care Automation
ADM+S Working Paper Series,
16 August 2024
RESEARCHERS
PARTNERS
Australian
Broadcasting
Corporation
Australian
Red Cross
Telstra
