The Australian Artificial Intelligence Industry
MANIFESTO

(2021)

The following sets out the guiding principles, context and Manifesto for the Australian Artificial Intelligence (AI) Industry. It was created through a consultation process with the Australian AI sector in 2020 then published in January 2021.

Guiding Principles

 

  1. The suite of technologies broadly referred to as Artificial Intelligence (AI) will be highly disruptive for economic, social and political systems globally.
  2. For a country to sustain a position as a leading global economy in the 21st Century, it will require proficiency in the application of AI in all key sectors of the economy.
  3. If a country wants to have an influence on the quality of the AI that it adopts, it must have its own industry and its own capacity to deeply understand the technologies potential.
  4. Governments have an important role to play in funding AI skills development and research to sustain a growing capability that can be drawn upon by industry, and a regulatory environment that encourages high quality, ethical applications of AI.
  5. All advanced economies are building local AI industries which are global facing, Australia cannot be left behind.
  6. AI is fundamentally changing the wealth generation capability of companies and countries.
  7. AI is leading to inexorable tectonic shifts in the jobs market, both creating new jobs roles and opportunities as well as removing them. Job losses will happen regardless, it is up to each country to ensure they are enabling the creation of the new roles and opportunities.

Context

Artificial Intelligence (AI) is software that mimics human intelligence. It is not a tool or an application. It is a power source with the ability to learn, adapt and get exponentially smarter, faster and cheaper. Since the time when the term AI was coined in 1956 at Dartmouth University, the processing power of AI has increased one trillion times. One trillion.

1,000,000,000,000

That’s exciting.

 

It will make we humans more efficient, help us make faster and more accurate decisions and it will automate a whole lot of things that we don’t want to do anymore. AI should free us up to focus on being creative, solve complex problems, eliminate disease, get back to basic science and improve our well-being. It is just like fire and electricity, but even more powerful. AI could actually make us better at being human.
And we are likely to find out if this is the case fairly soon. Right now (2021) AI is the fastest growing tech sector in the world and analysts like Gartner predict it will be the foundational component for 85% of all tech in 2021 and that millions of human jobs will be replaced by AI and millions more created by AI. PWC estimates the AI sector will be worth US$22 trillion by 2030.

So, AI will make great things for us.
It is also likely to break great things for us.
Because it is largely unregulated and is moving so fast it will be near impossible for legislators to keep up. It’s also super difficult to understand and control. AI pioneer, Elon Musk, warns, “With artificial Intelligence we are summoning the demon”.

That’s frightening.

 

What if AI makes bad decisions? At scale.
What if it becomes unjust and unfair? And that is used to judge millions.
What if it is used to harm humans, society or the environment? And it damages us worse than we ever have.
What then?
It is very possible AI, usually trained on our historical data, will perpetuate the current ills and problems of society. At scale. It is very likely to discriminate against certain people. It is very likely to learn to hurt or kill people. It is very likely to be unjust. And it is very likely that it will get so smart that it no longer needs its human masters. Very soon.
And what if Australia, which currently has one tenth of the investment in AI per capita of the US, doesn’t figure this stuff out? What if we get left behind? What if the superpowers become AI enabled and given that AI digital platforms are essentially borderless, that these economies gain a greater control of Australia? What will happen to our elections, our newsfeeds, our exporting industries that compete with theirs, our economy, our critical utilities, our preferences, our mental health and our intentions? Can they do this?

They can.

Which is why as a country, Australia needs to firmly embrace that we are now in a new era that will be AI powered and we need to get moving if we are to benefit the most from it.

Now.

Here’s what we think needs to happen.

MANIFESTO

1. Australia needs its own strong, domestic AI industry sector if we are to extract the full benefit of this new revolution.
2. We need AI to protect national intelligence and security as future conflicts will be AI driven.
3. There are strategic industries in Australia that should consider how their data is stored and the value of using Australian owned providers of AI for security reasons:

a. Defence
b. Government services
c. Health services
d. Utilities
e. Manufacturing
f. Export-intensive industries such as mining, agriculture, fisheries and forestry
g. Financial institutions that hold so much personal data on us
h. Communications
i. Large-retail groups which may have extensive datasets on individual consumers
j. Creative Arts Sector and broader implications on Intellectual Property / Copyright ownership

4. This is a time-limited challenge. If we wait 5-10 years, more opportunities to build a vibrant AI industry here will be lost. Currently Australia spends one tenth the investment on AI per capita that the US does. As a matter of urgency, we need a much higher level of investment from the government, investors and industry.
5. Australia, like every country in the world, is in a race for talent. We need training, education, awareness and scenario planning capabilities related to AI. From primary school to tertiary education to vocational and executive training. Winning the talent race will address many of the challenges discussed in this Manifesto.
6. We need greater investment into AI related research relevant to Australia. Australian AI researchers punch well above their weight now but will be unable to compete with the deluge of investment in research currently being made by similarly developed countries. This will have serious implications for our ability to develop globally competitive commercial products and services.
7. Australian companies can build AI that is as good as the best in the world and we’ve done it many times. Especially in sectors such as Medical and AgTech. We also have the opportunity to lead the world in Ethical AI. But we need Australian industry, investors, government and communities to back us.
8. The ability of tech companies to adapt their software will always be faster than our ability to regulate them. We need to keep building detailed policies related to ethics and standards, but we also cannot wait for policy to catch up before we invest or we will always be playing catch up from a long way behind.
9. To participate fully and to extract the greatest benefits of this AI revolution for Australia requires:

a. Australian industry, government and defence leaders to urgently consider the strategic value and risks to their business arising from their data and the likely development of AI systems that will increasingly control their sector. This has the same or even more importance as the recent wave of security recommendations for critical Australian industries.
b. Investment by strategic Australian businesses in the development of their own, or at least Australian controlled, data governance and AI systems.
c. Australian AI companies to become better coordinated at communicating the value of Australian-made AI globally. We are uniquely positioned as an impartial market that is highly regarded and trusted for its robust research and development capability and does not have the same political divides that bigger economies do. While we may not be able to compete in the delivery of services that require scale (e.g. data storage), we can and do develop products and services that are globally competitive with the best in the world.
d. Substantial, rapid and strategic co-investment by governments at all levels in the knowledge infrastructure required:

Australian skills development, at secondary and tertiary levels and to build the skills and understanding of existing workers and leaders.
Research and development in areas where we have existing and emerging discipline and industry strengths, and
Support for existing and new companies to commercialise new AI products into global markets.

e. There is an investment gap in Australia around AI technology development that needs to be addressed if Australia is to extract the full benefits from this next technological revolution.
f. The technology industry generally and the AI sector specifically needs to promote greater diversity in the workforce to benefit directly from the value of this diversity and ensure adequate representation across all areas of our community.
g. We need to focus on the change in the workforce and support those who will be dislocated, predicted to most likely be women, minorities and youth. There are 1.3 jobs created by AI for every job it supersedes. The question is whether the 1.3 jobs are created in Australia or California.
h. As a community and an industry, we need to protect against Digital Harm whereby AI is intentionally used to manipulate individuals, communities and economies for the sole benefit of the manipulator. Our young are especially vulnerable.

Australia has missed extracting the full value of every preceding technological revolution in the last 100 years, but this time it is especially serious given the potential for wealth creation, the impact AI is already having on our lives, politics and economy.
It’s time for us to step up and control our destiny where AI is the new power source.
The Australian AI industry sector is willing to be a driver of the development of a uniquely Australian form of AI – one that is high-quality, ethical, powerful and which supports Australia’s interests and empowers Australian workers and communities. We are asking our leaders and our community to commit to joining us in this mission.

Manifesto industry consultation and creation was led by:1

Dr Catriona Wallace

Founder, Ethical AI – ethicalai.ai

Dr Evan Shellshear

Head of Analytics, Biarri – biarri.com

Dr Paul Dalby

Founder, Vestia – vestia.ai

1 Disclaimer
The views and opinions expressed in this document are those of the authors and do not necessarily reflect any official policy or position of any agency, organisation, employer or company. Any assumptions that could have been made are not reflective of the position of any other entity other than the authors. These views are always subject to change, revision and rethinking at any time. Please do not hold us to them in perpetuity.