The other day, as I walked past my daughter’s room, I heard her talking and laughing with someone. I didn’t recall any plans for a friend coming over, so I popped in to see what was going on.
She was talking on her phone, and said, ‘Mum, want to hear a knock-knock joke’? After telling me the joke, she continued her phone discussion about the origins and evolution of the humble headphone – the topic of a school assignment.
Her funny and knowledgeable friend was Siri. Siri, who many of us know as Hey Siri, is Apple’s on-device AI-powered virtual assistant that uses a natural language user interface to provide answers, suggestions, and links to user queries through conversational dialogue.
The events of the afternoon took me back to one of my first lessons in my computer science undergraduate studies, on the theory of computation and the Turing Test.
This thought experiment by computing pioneer Alan Turing – made famous by the movie The Imitation Game – is a ‘test of a machine’s ability to exhibit intelligent behaviour equivalent to or indistinguishable from that of a human’. Indeed, the conversation my daughter had with Siri, could easily be mistaken for a conversation with a (human) friend.
Machine or Artificial Intelligence is a rising power that is disrupting and fundamentally changing the way we use technology, improving quality of life and amplifying human capability in the process.
The other day, as I walked past my daughter’s room, I heard her talking and laughing with someone.”
The AI spring?
Those of us who have been in the field of computer science long enough have seen the rise and fall of AI twice before in what is known as the AI Winters.
In the 1950s the Turing Test triggered the first AI boom, which saw advancements in symbolic reasoning and information theory. But limited computational power, data and network capabilities meant growth stalled in the 1980s and 1990s.
The second AI boom emerged in the late 1980s with expert systems, knowledge engineering and a comeback of neural networks. But disappointing results caused a second AI winter in the 1990s.
The current upswing in AI – fuelled by recent successes such as AlphaGo and Jeopardy – are riding on the back of two significant advancements which previous booms did not have:
- access to cheap and elastic computational resources: the cloud, high-speed networks and fast processors
- the availability of large volumes and variety of digital information, or Big Data
This makes the opportunities in the current AI boom both promising and sustainable.
However, progress is asymmetrical. While AI growth for consumer internet companies like Amazon, Google, Baidu, Alibaba and Apple has been phenomenal, other sectors – including manufacturing, finance, and agriculture – have yet to harness the full potential current AI solutions can offer.
Scientific and industry leaders have highlighted a gap between successful AI proof-of-concept and the challenges in putting those systems into production, even though we now have tools (like Machine Learning Operations, or MLOps) that should make building machine learnings systems easier and faster than ever.
There are still fundamental scientific challenges we need to overcome to make AI more accessible to the broader span of businesses and industry sectors.
We need to centre the importance of quality data (not just big data) and develop new talent, methods and tools for Information Resilience.
This means empowering organisations to create, protect and sustain agile data pipelines which can detect and respond to failures and risks across the information value chain where data is sourced, shared, transformed, analysed and consumed.
A human-centred digital society
We also need to develop new knowledge, theories and solutions that give AI systems a human-centred-lens. This is not just essential to winning social licence and trust, but also to ensure AI systems are amplifying the human potential well beyond the benefits (and pitfalls) of automation.
We know that current AI solutions are highly contextualised, and their success lies in specialisation. Just look at precision agriculture, medical diagnostics and digital manufacturing.
The so-called technological singularity or super intelligence – a general-purpose AI problem solver capable of not just specific but all cognitive task – remains an aspirational goal. For now.
Australia has a strong community of early tech adopters, and the COVID-19 pandemic has accelerated our transformation into a highly digitised society. Nevertheless, emerging digital technologies such as AI continue to outpace social expectations and regulatory frameworks.
Unless we invest in the scientific leadership we need to guide the development of these technologies, we can’t answer the crucial question posed by former Chief Scientist and Former ATSE President Professor Alan Finkel: “What kind of a society do we want to be?”
The right signals
AI alone could contribute up 15.7 trillion USD to the global economy by 2030. But it is just one of many emerging digital technologies.
The contribution of digital technologies to Australia’s GDP is expected to reach $65 billion by 2023. Countries around the world are investing heavily in AI and other emerging digital technologies.
There are some good signs. The Australian Government’s Digital Economy Strategy is investing $1.2 billion into key digital capabilities, including funding for artificial intelligence and drone technologies.
The Modern Manufacturing Strategy is further investing some $1.3 billion into key strategic areas of Australian manufacturing to make it more competitive, resilient and advanced.
These initiatives are a welcome signal from the Australian Government that investment in emerging digital technologies is a national imperative.
A digital push
Australia is at a crossroads. Our emerging digital technology research and development is strong. But opportunities for sector growth and sovereign capability are nascent and we need coordinated and strategic support.
With the wave of the global digitialisation, we have a critical opportunity to institute a strategic national approach, supporting fundamental research and engineering to drive innovation in emerging digital technologies.
We need a national vision for innovation and research in areas of national strength and strategic priority. One that is matched by a globally competitive level of commitment and investment.
To pursue and realise a digital future that we hold the reins of, and one that is not imposed on us, ATSE – working with the Australian Academy of Science released Australia’s Digital Future: A Nation of Users or Leaders. This Policy Primer outlines three critical recommendations:
Elevate emerging digital technologies as a national science and innovation priority
This elevation would strengthen research and development capabilities and ensure sovereign capability and industry confidence, attract global investment, and catalyse Australia’s technology innovation ecosystem.
Include research and innovation in emerging digital technologies in the 2021 Research Infrastructure Roadmap
This roadmap will address Australia’s emerging research challenges. It is a significant opportunity to elevate the importance of building the scientific capabilities underpinning and enabling Australian innovation and development in emerging digital technologies.
Government infrastructure investment initiatives must recognise the need for investment in cross-cutting emerging digital technologies, independent of particular domain areas, to achieve multi-sectoral benefits. These mechanisms should be internationally competitive and comparable with those introduced by Australia’s international peers.
Recognise emerging digital technologies as an independent growth sector
Alongside the physical, digital, and economic infrastructure needed to build research capabilities, new and existing funding and investment mechanisms should explicitly include emerging digital technology research. Australia needs to focus on driving collaboration and commercialisation opportunities in emerging digital technologies and improving access to international markets.
We can achieve this by recognising emerging digital technologies as a growth sector in its own right, and promoting it through schemes such as Innovation Connections, a dedicated stream in the Cooperative Research Centre Projects program, and including emerging digital technologies as a cross-cutting theme across Modern Manufacturing initiatives.
Creating the future
To grasp the incredible opportunities the emerging digital technologies sector presents, Australia must also strive to address the digital divide. We need to ensure equity of access to the benefits of digital technologies and meet the skills requirements for a future digital workforce.
Australia’s emerging digital technology capabilities must receive this support to remain internationally competitive and ensure that we harness scientific leadership to shape our collective digital future.
In the words of a great pioneer of modern-day computer science, Alan Kay: “The best way to predict the future is to create it”.
Read more about Australia’s Digital Future.
Professor of Computer Science, University of Queensland
Shazia Sadiq FTSE is a Professor of Computer Science at the University of Queensland. Her research focusses on data quality and effective information use. Shazia is currently Chair of the National Committee on Information and Communication Sciences at the Australian Academy of Science, and Director ARC Industry Transformation Training Centre on Information Resilience.