Fostering a successful and responsible AI industry for Australia
With even the Pope now talking about artificial intelligence (AI) and its impact on humans, how should Australia move forward to foster a successful and responsible industry? Australia National University AI expert Associate Professor Charles Gretton shares his thoughts.
Australia’s AI industry should aim to provide us with access to new technologies and ways of working, improve the global competitiveness of Australian industry and raise living standards across the Australian community. Only by achieving these goals will it constitute a global success story, and lift our nation - before others - in the climb to the future.
Several policy pathways could help foster an effective AI industry in Australia:
Professionalism. Professionalism in the supply and use of AI services is essential for competent and effective AI service delivery. Industry codes of conduct, recognition of qualifications, non-discriminatory accreditation, and registration according to merit-based criteria will contribute to professionalism. Appropriate insurances and assurances covering public liability, product liability and professional indemnity, should be developed for, and in consultation with, Australia’s AI industry and practitioners.
Sovereignty. The development of new technologies, ways of working, and the provision and consumption of novel services are needed to realise the potential benefits of AI. As a global services activity and in contrast to material goods trade, cross-border activity and transactions may be difficult to identify and regulate. The nature of AI will challenge traditional concepts of sovereignty and jurisdiction between, and within, nation states. Australia should engage with trading partners to ensure AI innovations afford the greatest benefit to all our people. Regulation to preserve a sovereign industry and restrain foreign (corporate and national) overreach should be developed deliberatively, carefully, calibrated, and in consultation with experts and stakeholders. Checks, balances and accountability should be foremost in deliberations.
Human capital. A responsible AI industry capable of delivering optimal community-wide benefits requires a society with knowledge, experience, training, and understanding (i.e., human capital) to recognise and leverage the benefits of AI. There is an urgent and pressing need to fast-track knowledge and competency across the workforce so that developments in law, regulation, business practices, and use by industries and consumers, deliver the greatest good to government, industry, and the community.
Educational infrastructure. Educational outcomes relevant to the take up and use of AI in Australia will require new computing resources and network infrastructure to support education services relevant to all phases of AI development, evaluation and deployment. New AI-ready infrastructure is required to support school curricula across years K-12, continuing with vocational, technical and further education, and non-research tertiary education.
Human sanctity. Many Australian industries have a substantial services component, with roles performed by people using historical information and communications technologies. Australia’s AI industry must ensure that ‘human-likeness’ afforded by AI is by default distinct from human capability. If not, the workforce and wider community would be disadvantaged and it would be a stumbling block to the effective use of AI.
Reach. Realising the full social and commercial benefits of applying AI across Australia will require service delivery to remote, rural and regional locations. Policy related to the provision of network, computing and data infrastructure must meet requirements for equitable access across communities, social groups and businesses of different sizes and types, in urban, rural and regional areas. Meeting these requirements will benefit the AI industry and the wider community.
Adjustment support. Implementing AI will disrupt existing technologies, employment and social and communal norms. Some businesses and population groups may suffer unreasonably. The consequences of workforce changes and disruption from AI development and adoption can be identified in advance so that appropriate adjustment support can be provided to facilitate an efficient and equitable transition to an AI-integrated future.
AI is a rapidly evolving field. Contemporary and emerging technologies, and means to engage with AI, are driving rapid evolution in the AI industry. To maintain and improve Australia’s engagement with the scientific and technological development of AI and the uptake of new AI technologies, Australian research into AI warrants measured public support proportional to the benefits. This may take the form of government-funded pure research or as public-private partnerships between government and industry. Public support should be targeted to bring the level of AI research and development undertaken by individuals and businesses up to the socially ‘optimal’ level. Effective and well-resourced R&D will position Australia with a vibrant, productive and responsible AI ecosystem in highly competitive international industries.
Well-known cyber-punk writer William Gibson famously suggested the ‘future’ is distributed unevenly. Australia isn’t demanding enough of the AI future – we’re on the backward side of the distribution. Our governments have a choice – get serious or get burned. The more agency we relinquish, the more we opt to follow as others lead, and the more likely we will find ourselves in a future that belongs to someone else, slipping ever further from our grasp.
A/Prof Charles Gretton convenes the ANU’s TechLauncher program
This article is drawn from the author’s contribution to the ANU’s IAI network submission to the Australian Senate Select Committee on Adopting AI.
