No longer science fiction, AI can personalise learning

Artificial intelligence promises to transform our economy and society, unlocking innovations and productivity that surpass the industrial and digital revolutions.

But Australia will need a much larger, highly skilled and globally competitive workforce to realise this productivity uplift.

It is reassuring, therefore, that the Australian Universities ­Accord released earlier this year sets ambitious targets to double the number of Australian university graduates by 2050.

The accord is full of recommendations to strengthen higher education and, importantly, increase the number of people with degrees from disadvantaged, First Nations and regional backgrounds. This will build a fairer and more productive Australia.

The accord also recognises the growing importance of lifelong learning to the reskilling and ­upskilling of workforces. However, it makes no recommendations to harness the transformative potential of AI to innovate personalised learning and to accelerate scientific breakthroughs with real world impact.

It is no longer science fiction to envision a living curriculum that is centred on each individual student. Despite our best efforts to innovate in teaching, learning ­remains largely a one-size-fits-all endeavour delivered to class ­cohorts.

But AI can deconstruct the classroom to a granular level of personalisation, where each learner can be assessed and supported and progress differently according to their abilities and aspirations.

For students from equity backgrounds (including students from non-English speaking backgrounds; those who have a disability; women in non-traditional areas; students who identify as indigenous; are from low socioeconomic status locations; and are from regional and remote locations) or who have learning support needs, this could be a game changer. AI will also give us more agility to break up degrees into bite-sized modules and empower students to design their own learning journeys.

AI will also help tackle the complex task of awarding credit for prior learning, including from work experience, enabling students to progress more quickly through their degrees.

More immediately, we must introduce AI literacy and technical skills into all our courses to prepare our graduates for future workplaces that have been redefined by AI.

Most workplaces are already innovating with AI, as revealed in a recent global survey from Microsoft that found 75 per cent of knowledge workers were using AI tools to boost creativity and individual productivity.

Similarly, most students are ­already using AI well before they step into the classroom. The first university to partner with ChatGPT creator OpenAI, Arizona State University in the United States, gives staff and students unlimited access to ChatGPT4, with safeguards for privacy and ­intellectual-property protection.

While human intellect and ingenuity will remain central to research, AI will accelerate scientific discovery. CSIRO reports “steep rates” of AI adoption in all fields of science, making research cheaper, faster and safer.

The AlphaFold AI program created by Google’s DeepMind solved a 50-year-old grand challenge in 2020, by surpassing human expertise in protein structure prediction, which then led to accelerated scientific discoveries for millions of researchers worldwide. Released a few weeks ago, its successor AlphaFold3 goes beyond proteins to predict the structure and interactions of almost every biological molecule known to ­science.

Universities account for roughly one-third of expenditure in national research and development. Thus, leveraging university research capabilities will be central to building a sovereign capability in AI.

The Universities Accord signals that our world-class research does not readily translate into sources of innovation for the needs of industry, government and society. Here too, AI will help. AI assistants will be able to match industry needs with academic expertise and translate academic research into more easily digestible formats for industry.

If we get it wrong, AI could increase inequalities and reduce human agency. Synchronous to AI innovations, we must address the escalating risks of AI, with issues such as its carbon footprint, grossly underpaid workers training AI, “prompt flooding”, “fauxtomation”, “softfakes” and numerous epistemic risks, requiring broad consultation and not just self-regulation.

AI adoption must be guided by strong ethical frameworks to enhance wellbeing, ensure accountability and protect data privacy. We must anticipate how AI will change our economies and societies, and put in place appropriate regulatory safeguards, invest in reskilling to support workforce redeployment, and mitigate the risks of unintended outcomes. With deep expertise in ethics, universities have much to contribute here.

Regional and rural Australians are particularly at risk of being left behind due to digital inequality and inaccessibility. Universities with regional strengths such as La Trobe have an important role to play in leading conversations about responsible AI adoption, while also providing AI literacy skills and training that uplift our general capabilities for working and living with AI.

Australian universities must refresh their education and research strategies to drive an AI-first transformation that delivers equity to the higher education sector and innovation across the national economy.

Universities that are deeply embedded in the regional communities of Australia are ideally positioned to lead this sector-wide uplift for participation parity and workforce productivity.