Can Australia lead the AI race?
Distinguished Professor Mary-Anne Williams discusses risks and challenges facing Australia in an AI future.
“AI is changing our lives, businesses, institutions and society,” says Mary-Anne Williams, Director of Magic Lab in the UTS Centre of Artificial Intelligence. The Distinguished Professor and her team are bringing science fiction to reality with research on AI, Social Robotics—machines that safely interact with people—and their future ramifications.
The changes will be profound: “A new world order, similar to how electricity dramatically changed civilisation. What industry has not been transformed by electricity?”
Australia’s place in this AI-enabled future is the topic of her UTS Distinguished Lecture Series talk, Artificial Intelligence Supremacy: Will Australia lead or lose in a global AI World.
Williams’ interest in artificial intelligence stemmed from a young age. “I grew up in a family business that relied on technology and innovation for competitive advantage,” she says.
I was curious to understand [AI’s] capability for solving practical problems and shedding insights on how the human brain works.
When introduced to computers at university in the 80’s, her immediate thought was their application to monitor stock and order inventory more efficiently than paper records. She subsequently built an intelligent database management system for her first job at the Co-op book shop in 1985. “I used AI to predict book sales; these predictions were used to make more informed decisions for book orders and new releases.”
Williams believes Australia is well prepared for AI leadership in terms of significant investment in digitisation and digital transformation. But that’s just laying the groundwork; what’s missing is a strategic course and vision.
Her talk outlines four key challenges: “Develop governance and policy frameworks that maximise benefits and minimise risks of AI; ensure that AI is not used to abuse power and reduce access to its benefits; develop responsible AI technologies that are safe and secure; and develop a competitive market in AI education and training.”
It’s essential that all areas are addressed simultaneously to realise the promised productivity gains, Williams asserts. For starters, government can help resolve current skill bottlenecks and tensions with industry demands for AI talent. It also needs to provide incentives to develop AI technology, particularly in startups, and boost adoption to make it more scalable and productive.
All nations face the same challenges; Australia can lead because it has the opportunity to address all four challenges better than larger, more complex, fragmented and politically polarised countries.
Bold actions are needed on funding. With AI a national priority in the US and China, Williams says the former’s Defense Advanced Research Agency (DARPA) is committing $200 billion over five years. “We can learn from the leaders.” Accounting for population differences, “Australia should spend at least one percent of DARPA funding on ensuring we lead in AI—that's two billion over five years.”
Her talk also highlights the depth of transdisciplinary and diverse AI talent in the Magic Lab, which includes extraordinary leaders and mentors such as Apple Co-founder Steve Wozniak; Peter Gärdenfors, who served many years on the Nobel Prize Committee for Economics; and Henri Prade, UTS's most highly cited researcher.
Williams feels privileged to work with the country’s best and brightest. “UTS is currently Australia’s undisputed and uncontested leader in AI and Social Robotics research.” She says the field of social robotics is young and expected to impact a wide range of industries. Examples include robots as IoT gateways, companions or carers in the home; workplace assistants; and for customer experience in business.
Her long-term vision for Magic Lab involves consolidating its current position in Social Robotics in Australia, and aiming for global leadership by scaling and enhancing research capabilities toward a critical mass of expertise. Other goals include pursuing new research and funding opportunities to support students and early career researchers, and developing strong, trusted and sustained partnerships with industry and government.
AI is difficult to understand and its impact even more difficult to predict. But we cannot ignore it.