Wake Up Australia

Our inaction on artificial intelligence is costing us now, and disadvantaging us for the future.

If there was an alarm clock for national prioritisation and investment in Artificial Intelligence, Canada surely set it for 6 am (being the first nation to launch a national AI strategy in 2017). This spawned rapid, vast and significant global activity (or at the very least, formalised it). Numerous organisations, institutions, and local, state and federal governments across the world have prioritised AI, and taken significant action (and allocated substantial funds) to address the challenges and opportunities it presents.

AI advances and applications are being applied and adopted by industry at an accelerating pace, and across ever-increasing domains. From agriculture and healthcare, to space exploration and transportation, AI is rapidly expanding our capabilities and changing the competitive landscape. By 2030 it is anticipated AI will add US$15.7 trillion (or nearly 12 Australian economies) to the global economy. AI will create labour market displacement, economic shifts, and disrupt varying domains. It will significantly impact the lives of nearly every person directly or indirectly, with both positive and negative effects. The way the benefits and advances will be accrued and distributed, however, remains to be seen.

How nations address both the challenges and opportunities of AI will, in large part, determine the quality of life for citizens in the future, as well as the future global power dynamic. Given its expected impact, AI represents a strategic imperative for the economic, social, political and environmental sustainability, security and prosperity of all nations (and their citizens) across the globe. It is partly for these reasons many national AI strategies attempt not only to understand how to regulate or adapt governance to this advanced technology, but endeavour to rapidly and purposefully develop sovereign AI capability.

A review of the global AI landscape identifies three key insights:

1) There is a large and increasing level of AI prioritisation, investment and activity underway globally, spanning organisations, academia, and government.

2) National AI strategies take unique positions, are increasingly detailed, and vary in mechanisms and objectives; yet have common threads.

3) Australia’s footprint on the global landscape is limited, reflecting our lack of national prioritisation and commitment to AI, posing both immediate and long-term implications and consequences.

Given these three insights, it is clear Australia requires a national AI strategy, and must take appropriate actions to meet the demands and challenges of the 21st Century. Such strategy must be cognisant of the ever-changing global landscape, tailored to our unique position and possibilities, and reflect a broader shift in economic and social thinking. It should also be clear that Australia requires this AI strategy and shift in prioritisation and thinking immediately, as we are already falling behind and there is no overnight solution.

Global AI Activity is Established and Accelerating.

China, Finland, Taiwan, Luxembourg, South Korea, Singapore, India, France, Germany, Japan, Estonia, and the US, UK and European Union are just some nations and multilateral institutions which have crafted (and to varying degrees) implemented AI strategies. Other nations are rapidly following suit, with Russia, Malta, Kenya, and others all announcing, or proposing, the development of AI strategies. The European Union has further proposed all member-states develop national AI strategies by the end of 2019. Alongside the countless mechanisms and actions proposed and detailed in each strategy, nations have committed millions, billions and even tens of billions of dollars toward AI (detailed below) advancement. Steps taken by nations are not limited to domestic action, but increasingly reach the international stage, where nations are engaging in discussions to shape global approaches to AI. For example, Canada and France have created a new expert International Panel on Artificial Intelligence, and AI has become a key topic at both the OECD and G7.

Many companies are dedicating vast resources to AI, with Amazon, Google, Facebook, Alibaba, Tencent, Microsoft, IBM, Intel, NVIDIA and many others committing billions of dollars, and vast resources, to AI research and development. Universities are committing to AI, with MIT, Harvard, the University of Toronto, and Oxford University being just a handful across the globe committing hundreds of millions (if not billions) of dollars, and talent, toward AI advancement and management. For example, in 2018 MIT announced it is creating a $1.38 billion AI centre. Both university-housed and independent institutions have been established to advance AI research, development, regulation and management. Open.AI, The Future of Life Institute, AI Now, the Stanford Institute for Human-Centered AI, and the Machine Intelligence Research Institute are just some of the many institutions focussed on AI.

Disruption to industries, sectors and entire economies is underway, with new AI based companies being borne, and increasing numbers of AI enabled companies, changing current approaches every day. AI startups are growing, and are becoming a key element of national capability, innovation, and bottom-up adoption. Venture Capital investment toward AI is steadily growing, with global VC investment in AI reaching $13.2 billion in 2018 (up 120% from 2016).

This combined action and effort by industry, academia and governments is accelerating the pace of change, as well as the size and scope of potential disruption and advancement globally. It is also a driving force of, and response to, national AI strategies and prioritisation.

National AI Strategies Vary, But Address Common Themes.

Many national strategies are borne from a realisation of AI’s economic and social value, as well as strategic importance. Such adages are reflected in the strategies, objectives and comments on AI by key nations. The UK Select Committee on Artificial Intelligence noted “the UK can either choose to actively define a realistic role for itself with respect to AI; or be relegated to the role of passive observer”. This sentiment has been echoed in other national approaches, with China pledging to become the world leader by 2030 and the US placing AI as one of its highest national R&D priorities (behind national security). The Finnish interim national strategy also notes the importance of a realistic and coherent appraisal of the nation’s capability and potential, warning against the threat of the nation “lacking the courage to engage in large-scale reforms”, “slow and ineffective commercialisation”, and the weakness of “a culture of avoiding risks”. Taiwan’s national strategy seeks to “create a sound ecosystem which creates a niche market, and ensures Taiwan remains an important partner in the global intelligence system value chains”. Germany, France, Singapore and a host of other nations are all vying to become global leaders in the advancement, development and implementation of AI technologies, within a rapidly advancing and changing global landscape.

While each national strategy takes a unique approach, leveraging different mechanisms and reforms, a general focus on skills and talent, research and development, commercialisation and adoption, as well as managing the technologies effects, can be found in each. These mechanisms and strategic actions seek to amplify enablers and drivers, remove blockers, minimise any potential negative consequence, and enable a nation to achieve a stipulated objective or position (such as becoming a global leader in the application of AI technologies). While not exhaustive, nor wholly inclusive, the below section seeks to highlight the breadth, depth and scale of action taken, and prioritisation given, to making AI a key sovereign capability throughout the world.

Skills and Talent

National AI strategies generally seek to achieve two objectives when it comes to a focus on skills and talent; first, develop a national AI workforce; and second, provide support to those impacted by labour market displacement resulting from innovation and disruption. A national AI workforce is critical to a nation’s capacity to research, advance and implement AI applications within government, academia and industry. Competition for talent is intensifying, with salaries for world leading experts reaching the hundreds of thousands and even millions of dollars, straining the ability of nations to develop and maintain an AI workforce. Recent analysis by Element.AI (a leading Canada-based AI company) has detailed the current global AI workforce, noting the shortage in global talent (despite recent growth).

Responding to intense competition for talent, nations are implementing strategies to develop, attract and retain AI expertise. China’s approach to AI reportedly calls for the nation to develop 100 “AI + X” cross-disciplinary majors by 2020, have 50 national online open courses of a high-quality, and has launched a program to train 500 teachers and 5,000 students in AI-related skills over the next five years. The UK is seeking to augment strong expertise by offering an additional 1,000 PhDs, doubling the number of high-expert visas to 2,000, and also calls for AI to be added to the immigration Tier 2 shortage list. It will also provide funding and develop a prestigious and global Turing fellowship which will offer studentships (valued at $173 million) to increase domestic talent, as part of its approach. France will double the number of students trained in AI, and establish a national AI programme. India’s strategy calls for the nation to develop initiatives such as expanding education in AI and data science (including credit-bearing MOOCs), promoting employee re-skilling, introducing AI in schools, and by developing AI training standards. Nations including Canada, France, Taiwan and the UK are also focussing on the ability to attract and retain global AI expertise, by prioritising the skill-set, shortening the immigration processing requirements, and offering a more favourable environment and competitive industry.

Taiwan aims to cultivate 1,000 elites in intelligent technologies by 2021, train 10,000 pioneers in intelligent applications, and seeks to recruit global AI expertise by offering innovation clusters, expanding existing programs, and by developing a new act for the recruitment and employment of foreign professionals. Singapore will implement several initiatives to develop domestic capability, including offering AI education across primary, secondary and tertiary levels, as well as for professionals and the general public. For example, Singapore will provide comprehensive workshops, adapt curriculums, and offer blended training programs as well as an AI apprenticeship programme. South Korea reportedly plans to establish six new AI graduate schools to increase the number of specialists, create 4,500 AI scholarships, and provide six-months of intensive training to 600 talented youth by 2021.

Nations are also taking steps to improve digital literacy and general AI awareness of working professionals to help prepare citizens for AI adoption, labour market displacement, and to improve general aptitude. Singapore’s Skills Future initiative aims to improve digital awareness and literacy amongst all Singaporean adults, including those planning to return to the workforce, by developing the appropriate mindset and basic functional skills required in the digital economy. Finland is aiming to guarantee AI literacy for all citizens, and has partnered with the University of Helsinki to offer a free 30-hour MOOC to learn the basics of AI, which has already had 90,000 individuals sign up. The French Villani report proposed the mass training of civil servants in digital mediation, and in the United States, Harvard University is providing AI education to elected members of congress. The United Kingdom has recently announced a National Retraining Scheme to prepare adults for changes to the economy, including those brought about by automation, aiming to help those affected retrain into better jobs.

In an age of intensified competition for AI talent, and with growing shortages, and in an era where digital awareness and digital skills are of increasing importance to the competitiveness and maturity of an economy, it is clear many nations have already prioritised and taken action to ensure they can develop, attract and retain the required talent to succeed.

Research and Development

As a core element of AI capability, the ability to undertake fundamental and applied research, and to advance technical capabilities, will help nations derive technological advantage and deliver breakthrough innovation.

As part of its announcement, the UK will invest $160 million from the Industrial Strategy Challenge Fund into the robotics and AI in extreme environments program (for use in industries such as deep mining, offshore and nuclear energy, and space), and has nominated the Turing Institute as the national AI research body. The UK research group EPSRC has also committed to providing $519 million for funding research related to ‘Data Science and AI’. Canada’s initial strategy was crafted to establish interconnected nodes of scientific excellence in the nation’s three major centres for AI (Edmonton, Montreal and Toronto), and to support the national AI research community. Finland will create an AI accelerator and Centre of Excellence for AI Research, South Korea will launch an AI R&D challenge similar to that of DARPA, and Japan will form three primary research centres to coordinate AI R&D objectives. India has proposed creating new Centres of Research Excellence in AI (COREs) which will focus on fundamental research, as well as International Centres for Transformational AI, which will focus on applications in key domains. Germany will make AI a priority for an envisaged Agency for Breakthrough Innovations, support AI advancement across numerous industries and sectors, and support research on explainability and accountability of forecasting and decision-making systems. Germany will also develop existing, and create additional, Centres of Excellence for AI at supra-regional levels, and develop them into a national network of at least 12 centres and application hubs.

France will create individual AI research chairs to attract researchers, and will further make calls for proposals which attract the best research projects. In addition to academia and government backed AI institutes, many nations such as Taiwan, will actively seek to attract global firms to establish AI research centres in their nation. The US National AI R&D strategy calls for a focus on long-term investments in AI research as a core element of the nation’s approach to AI, investing and supporting not only fundamental research for AI approaches, but in advancing hardware for, and with, AI. To increase AI R&D China’s proposals include the establishment of a $2.78 billion Beijing AI Park and support for national AI champions, such as Alibaba, who will invest $20.8 billion toward research and development heavily focussed on AI over the upcoming three years.

This focus, support and commitment to AI research and development is driving both current and future advancements, and is likely to help drive future advantages and collaborative developments.

Commercialisation and Adoption

Where AI research will help drive the future of AI capability and technical breakthroughs, the benefit gained will be limited if nations are unable to effectively and appropriately commercialise and adopt the technologies to transform their economy and society. Aiming to drive bottom-up and top-down adoption of AI, many strategies focus on AI startups, public-private partnerships, AI challenges, regulation adaptation and standard setting, data accessibility and security, the creation of testing arena’s, and creating AI ecosystems.

The US National AI R&D strategy identifies several key strategic priorities to enable the development and adoption of AI, including developing shared public datasets, environments for AI training and testing, measuring and evaluating AI technologies via standards and benchmarks, and by expanding public-private partnerships to accelerate advances in AI. To improve adoption China is creating a national AI park in Beijing (with an expected 400 enterprises and target annual output of $10 billion), the city of Tianjin has reportedly announced it will provide funds worth $21 billion to support the AI industry, and the nation is working to establish a foundation for AI industry support systems (such as establishing an open, high-quality annotated data resource base with standard test data sets). China’s strategy calls for AI to be a key source of growth by 2020 (with an AI industry exceeding $29.7 billion), a primary driver of industrial advances and economic transformation by 2025 (worth $79 billion), and by 2030 for the nation to be the world’s top AI innovation centre (exceeding $197 billion).

To enable the continued development and adoption of AI, the UK will take several steps to further strengthen critical infrastructure and supporting systems. This includes publishing more high-quality public data in open, easily findable and reusable formats; developing equitable, fair and secure data sharing frameworks; and by delivering strong digital and telecommunications infrastructure across the nation. Building on its existing strategy and strong AI ecosystem, the federal Canadian government has identified and supported five key innovation clusters throughout the nation, with a heavy focus on AI. One specific AI supercluster, focussed on intelligent supply chains, has received nearly $237 million in federal government funding (with further government and private funding contributed), with other superclusters receiving similar (but slightly less) funding. India has proposed experimenting with a National AI Marketplace Model, establishing an AI database on projects, and by directly supporting AI startups, as well as encouraging collaboration to implement the “AI + X” paradigm.

Taiwan will develop international AI innovation clusters, foster 100 AI-related startups, and create an international innovation hub (providing R&D facilities and test fields), with the objective of developing 100+ AI solutions. Taiwan will also seek to discover advantages by adopting models similar to DARPA, with topics specified by an “Expert Review Board on Scientific Research Plans”. South Korea announced it will initiative large-scale AI projects in the public sector, undertake long-term high-risk projects, and take steps to support local startups and SMEs; by creating an AI-oriented startup incubator, making a new supercomputer available to startups and SMEs, and by investing an additional $1.24 billion toward the development of an AI superconductor by 2029. To focus their efforts, Singapore will launch a series of AI Grand Challenges to promote bold ideas and innovative technologies, initially focussing on Health, Finance and Urban Solutions. France will aim to strengthen synergies between public research and industry, develop an open data policy to spur activity, support the creation of a public and private data exchange, create a regulatory and financial framework which favours the emergence of “AI Champions”, and seek to create a “European DARPA”.

Managing the effects of AI

Many national and multinational strategies also seek to identify and resolve potential challenges and areas of concern regarding the development and deployment of AI systems. From regulatory changes to developing standards and ethical frameworks, nations are endeavouring to ensure the advancement of AI is aligned with societal and ethical norms. Germany, which has already unveiled a detailed ethical plan for autonomous vehicle research, is reported to state AI must be embedded within a framework which protects fundamental social values and human rights. To address such concerns, Singapore has put forward three initiatives for the governance of AI; first, an Advisory Council on the Ethical Use of AI and Data; second, a paper on the responsible development of the technology by the Personal Data Protection Commission; and finally, a five-year research programme on the governance of AI and Data. Focussing on the international aspects of AI, Japan prepared AI R&D guidelines for discussion at both the OECD and G7. The European Commission formed a High-level Expert Group to advise on the implementation of AI within Europe, with particular focus on ethical, legal and social issues relating to AI. The HLEG has put forward Ethical Guidelines on AI as well as Policy and Investment Recommendations.

France will support human science research on the ethics of AI, make all algorithms used by the state public, and further encourage openness. The UK will launch a Centre for Data Ethics and Innovation, and consult with a new AI council, and appoint an Office for AI within government. The US will focus on both understanding the ethical, legal and social implications of AI, as well as measuring and evaluating AI technologies through standards and benchmarks. This will involve establishing AI benchmarks, developing AI testbeds, forming specific standards and by improving FAT (fair, accountable and transparent) AI by design. Canada has crafted a series of guiding principles to ensure the effective and ethical use of AI, developed an Algorithmic Impact Assessment, and has launched a Directive on Automated Decision Making. Australia is currently seeking feedback on a series of AI standards developed by Standards Australia, and the government recently sought input on an ethics framework for AI.

AI Investment

Beyond the announcement of priorities, mechanisms and objectives, many nations have committed vast sums of funding (from public, academic and private sources) toward AI advancement, development, deployment and management. For example, while the initial Canadian AI strategy provided $128 million, the federal government has committed $237 million to an AI supercluster (and further funding to superclusters which will likely leverage AI, totalling a billion dollars), the province of Quebec has put forward $223 million to its AI strategy, the Canada First Research Excellence Fund has invested $97.5 million toward Deep Learning and Optimisation, and Alberta will reportedly provide $110 million over five years toward growing the AI sector.

Japan’s 2018 AI budget totalled $1 billion (a 30% increase on the prior year), Singapore’s five-year funding is at $151 million, Taiwan’s annual budget for the Taiwan AI action plan was announced at $448 million (per year 2018–2021), and France will allocate $2.4 billion toward AI. In China, the Beijing AI Park has received $2.9 billion, and Shanghai will match Tianjin’s AI development fund of $21.76 billion. While the US strategy does not provide a federal number on investment, the nation had $8.3 billion in AI VC investment in 2017, unclassified AI R&D by the federal government reached $2.19 billion in the same year (up 40% from 2015), MIT will create a $1.38 billion AI centre, DARPA is investing $2.8 billion in AI, and the National Science Foundation provides over $146 million each year in support of AI research, and private investment is estimated in the billions.

Germany has announced its government investment of $4.71 billion is expected to be met with an equal $4.71 billion in private funding. South Korea’s national AI strategy was supported with $2.71 billion, the UK announced $1.7 billion in its AI Sector Deal, and the EU Commission has committed $2.4 billion to 2020, proposed at least $11.3 billion in investment over the 2021–27 budget, and aims to drive EU wide investment to $32.3 billion per year through the following decade. Global private investment is also increasing, with SoftBank recently announcing an investment fund with numerous partners (referred to as a ‘Vision Fund’) valued at approximately $156 billion, to invest in technology with particular focus on AI. While this is a snapshot and fraction of investment in AI globally, it clearly demonstrates the level of activity, commitment and resources dedicated to AI.

While the above section is far from a comprehensive analysis of the varying national AI strategies, nor their respective priorities and mechanisms, it should emphasise the vast amount of prioritisation and activity taken by many nations throughout the world. It should highlight the pursuit of AI and transformation toward a digital society is not only an issue to consider for the future, but is already well underway. If the advantages of tomorrow are being borne today, it is clear nations with AI strategies, who are actively and enthusiastically pursuing AI as a core element of technological, economic and social advancement, will bear advantage over nations who are not, both now, and increasingly, into the future.

Australia is falling behind in the global landscape.

Given the critical implications of Artificial Intelligence, and our relatively advanced economy and society, it would be fair to assume Australia is actively pursuing AI and aggressively transforming our economy and society, even becoming a leader in AI. Yet Australia, notably at the federal level, has failed to follow suit. While the 2018 budget provided limited funding ($29.9 million over four years) to craft a national AI ‘Roadmap’ (and other activities including an Ethics Framework), the national roadmap/strategy is yet to be released, and it remains unclear what, if any, actions and/or funding will be announced. The release of the 2019/20 Australian federal budget was another blow to those who had hoped (or perhaps expected), the Australian government would finally take note of the strategic importance of AI for the nation’s prosperity. But again, the government seemed to hit the snooze button on any AI prioritisation or investment (or have they simply unplugged the alarm?). In an age of increasing AI activity, competition and disruption, the resulting shift toward an AI “future” will largely determine the economic leaders of the future, as well as quality of life for nation’s citizens. It is within this context, we must evaluate Australia’s current standing and approach, and realise it is currently insufficient.

This, of course, is not to say Australia is devoid of AI capability, progress or areas of current or potential advantage, but there are immediate costs and long-term consequences and implications in failing to take increased action to prepare Australia (and Australians) for an AI future, and in developing sovereign capability.

Australia has key capabilities, advantages and has seen recent improvements, but not at the scale, scope or pace necessary.

In a positive step, the nation’s scientific research organisation CSIRO, a domestic leader in AI, has announced it will invest $19 million toward AI. Yet in a sign of current attitudes and level of commitment domestically, the organisation has continually highlighted the lacklustre state of AI (and innovation more broadly) in Australia, and need for the nation to pursue AI (and other) technologies with a sense of urgency. CSIRO Data61 was contracted to develop both a National Artificial Intelligence Roadmap and Ethics Framework for Australia (with the ethics framework released for consultation here). The organisation has also worked with varying organisation to develop and implement AI products and services across a variety of industries, and is currently a leader in fundamental and applied research. Government grants for research have also provided some support to AI capability over preceding years.

In recent years a series of academic and other AI labs and institutes have been formed. These include the Adelaide Institute for Machine Learning (Adelaide University), the AiLab, Centre for Artificial Intelligence (University of Technology Sydney), AI and Autonomy Lab (University of Melbourne), Applied AI Institute (Deakin University) and 3Ai (Australian National University).

While these academic labs are clearly beneficial and a core component of AI capability, and are driving the development of AI in Australia; hubs and organisations which focus on commercialisation, adoption and market innovation, will be needed for effective transition and implementation. AI research and advancement is an area where Australia currently outperforms our relative size on the global stage, frequently seeing top papers submitted and presented at leading AI conferences across the globe, as well as cutting-edge applications and techniques. We are also fortunate to have a strong base of talent from which we can build. We have also begun engaging the international community, notably via the Australian Standards Institute, which has been actively involved in crafting draft standards for AI (accessible here). Through our membership in the OECD we are also engaged in international deliberations.

Yet commercialisation and adoption, talent and skills, level of AI activity, awareness and commitment, as well as associated reforms remain an ongoing challenge, as does very limited investment levels. A recent report by Deloitte, for example, reported Australia has the second highest AI talent skills gap . Domestic AI startups and larger technology companies are helping drive AI capability and innovation, but more needs to be done. While Australia is home to cutting-edge AI startups, across varying industries and locations, the reality is the number of such AI startups is too small and restricted to become the key capability needed. More must also be done to create, grow and support the broader AI ecosystem. To realise our full potential, we need to increase the level of AI talent, and accelerate AI development and adoption, while ensuring we do so within the social norms and legal frameworks of Australia. Much more must be done to broaden the discussion and engage the general public on the capabilities, effects and implications of AI now and into the future.

Further analysis is required, as we need to identify our pillars of AI capability, develop and build foundations, and take stock of our strengths and weaknesses in relation to the global landscape to identify a future path, and ultimately position, for Australia in the global landscape. Despite having had AI capability in Australia for many years, if not decades, the reality is that in the new “race for AI capability”, especially compared to global actors, we are falling behind and more must be done. If we do not, we are likely to see a continued exacerbation of current hamstringing issues such as talent availability, see reduced responsiveness to change, as well as an inability to create disruption and achieve strategic advantage.

We need to prepare for AI now.

AI can have a transformative and beneficial impact for Australians. It may improve our health, reduce pollution and address climate change, it could enable us to spend more time on things we enjoy, it may reduce workplace accidents and broaden our knowledge and understanding of the universe. The possibilities are vast as AI offers the ability to affect nearly every aspect of our lives.

AI and other digital technologies are widening the disruptive capacity of companies and nations, and as such Australia must take a more global view of the competitive landscape, as barriers to entry lower, and global competitors increasingly compete in the Australian market. Yet awareness of increasingly disruptive and competitive markets is only a necessary first step. To compete in an AI-enabled world, we must not only react to disruption and seek to limit strategic surprise, but take steps to create it.

While it may appear we can merely pursue AI as a convenience, or ‘nice to have’ at our own leisure, others are pursuing the technology and its capability out of necessity (for example, the medical industry in China). These nations are unlikely to only disrupt and impact their own economies, but will actively seek international opportunities (placing additional competitive pressure on Australian companies).

Without the requisite sovereign capability, we risk strategic surprise and become vulnerable to critical dependencies on external vendors, and limit our capability to adopt many transformational applications. Developing sovereign capability is the best way to ensure we are in control of the path we decide to take.

Without proper engagement, guidance and action to address AI and its effects (such as labour market disruption, computational propaganda, or privacy concerns), we risk unintended or undesired social and political consequences. It is an important issue which requires action and discussion not just at the level of government, or C-Suite, but one which broadens the engagement to the general public. It should be clear AI is, and will continue to be, a defining capability and disruptive force economically, socially, and politically for decades to come. It should be even more clear Australia needs a vision, strategy and commitment to AI, and we need it now.

Addressing Australia’s AI Needs.

We must not only devise and implement initiatives designed to update current frameworks, but we need also engage in longer, more structural reform, to identify new foundations and approaches to a more futuristic, and digitally enabled, economy and society. For example, while we definitely need more investment in AI, and funding to support an AI ecosystem, we also need new channels to distribute such funding; with new criteria which spreads opportunities, improves and accelerates accessibility, and builds end-to-end (top down and bottom-up) capability and advancement.

Key to Australian success will be not only the capability to advance and develop technical AI (where we currently outperform our relative size), but in how we adopt and integrate the technology and its applications within, and across, industries, government and academia. It is also important to note it is not merely the technology itself which disrupts current approaches and creates new areas of competitive advantage, but the associated redesign of processes, offerings and strategies aligned with these enhanced capabilities.

We can, and should, leverage AI to solidify areas where we currently hold advantage, but simultaneously develop excellence in industries and areas of importance for the future. These areas of excellence and focus should not only deliver domestic benefit and prosperity, but offer vast export potential to global markets, and have a beneficial impact on humanity at large (i.e. healthcare). We should also continue to carefully manage its effects and amplify discussions concerning privacy, bias and other matters.

We also face challenges which are common internationally, and have a role to play on the international stage in defining the frameworks, regulations and norms in which AI currently does, and will continue to, operate. We should build off work currently underway, such as that by Standards Australia, and engage our global counterparts in discussions and ideas on how we proceed.

While this article is not intended to define and propose specific mechanisms or policies an Australian approach should include, there are some key areas where action should be taken. We must take steps to develop, attract and retain AI expertise to grow our domestic capability, and we should work to upskill and increase the digital aptitude of all Australians. We need to find ways to improve commercialisation to ensure the breakthroughs and knowledge we gain in research is translated into products and services adopted in our economy and society. We should work to rapidly accelerate the rate of adoption and transformation of our economy, focussing on both bottom-up (by supporting startups) and top-down (by incentivising public-private partnerships) adoption. We should develop and support Australian AI Champions.

To ensure its effectiveness we should review and update critical infrastructure and facilitate standards and best practice surrounding data. Our strategy should have as a core element discussions and proposals on the issues surrounding AI, in particular those such as privacy, bias and accountability. We should directly increase funding and investment to AI, and find ways to improve the process for accessing it, and improve the environment for those offering funding to support our ecosystem. We should develop tools, metrics and policies to measure, track and evaluate our AI industry and capability. We should also seek to spread awareness, knowledge and understanding of AI, and its potential impacts, to as many Australians as possible (from the general public to elected ministers). We should seek to accelerate projects which deliver clear and tangible benefits to the lives of all Australians. Our goal should be in line with our potential capabilities and with clarity of the global AI landscape, enabling Australia to develop areas of advantage and strategic positioning. We need to build the foundations for today and tomorrow, to ensure the resilience and adaptability of our approach, and take a realistic yet ambitious view of AI. If we manage to achieve the above, Australia will be in good stead.

Conclusion

Australia faces a choice; continue to try and protect and hold onto the past while resisting the forces of change, or embrace and prepare for the future. If the answer is (hopefully) the latter, we need to prioritise AI now, and take the much-needed steps, and provide the much-needed funding, to develop sovereign capability and excellence.

To do so, we must shift the dial from empty rhetoric to tangible action with clearly identifiable and measurable objectives and outcomes. Our vision must be bigger, bolder, and more ambitious than it is today. It should be a long-term vision, one which embraces the technologies and industries of the 21st Century, and provides not only economic growth, but broad and shared social prosperity. It must be embedded with a common purpose and understanding shared by Australians. It should shift our baseline for disruption, risk and the pace of change. We must accept this new reality, and see it not negatively as a threat, but as a positive and an opportunity — but this can only occur if we shift our perspective and take appropriate steps to adequately prepare ourselves. We cannot afford to wait any longer to do so.

To adopt a quote Otto Von Bismarck “if you are going to go through a revolution, you would rather undertake it than undergo it”.

Australia can, and should, undertake and drive this revolution.

We only need to wake up, and choose to do so.

By Michael Evans

Notes:

1. I was a key author of the yet to be released National AI Roadmap.
2. All figures are in AUD and converted using market exchange rates, unless stated otherwise.
3. If you have any information you believe should be added to the article, or corrected, please get in touch.