Australia in the wilderness on AI

milad-fakurian-58Z17lnVS4U-unsplash

The frenzy of discourse on generative AI had drawn focus from decades of work on other AI projects conducted without large language models, and seemed to have obscured the fact that Australia had underinvested in the area, the Chief Scientist of NSW Hugh Durrant-Whyte said yesterday.

Speaking at the Universities Australia conference yesterday, Professor Durrant-Whyte said he would present a contrarian view to others, but one informed by decades of work in the AI space.

“Australia is a bit of a basket case (in relation to AI), there is no money for anything other than regulation,” Professor Durrant-Whyte said.

Professor Durrant-Whyte said he had once worked with an organisation that employed 200 people working with AI, but now every one of them had left to jobs overseas.

“If I think of the top 100 people in AI in the world, none of them are in Australia.

“There is a whole area of AI where people are doing great work with AI which are making a profound impact of what we are doing in research and teaching but they are not large language models.

“By the end of this year we will fall off a cliff with LLMs and people will realise what they are good and are not good for.”

Chief Technologist at Google Australia, Natalie Piucco, presented a more upbeat view of the future of AI for universities – saying that they were already sitting on ‘a goldmine’ of world-class research that tech companies wanted to emulate.

Ms Piucco three key things that successful AI companies were doing to achieve positive outcomes in AI, that universities could learn from were:

  1. Partnerships – working with six partners to develop a successful AI solution
  2. Training for staff – making sure all staff had at least a baseline understanding of AI
  3. No Committees – getting a flexible operational structure in place to speed up progress.

Personalisation of technology was critical for business and technology success – re-thinking what student-centric technology really looked like, Ms Piucco said.

“If users can’t find what they are looking for in Google search, it is our problem. If they can’t spell, it’s our problem,” she said.

Share:

Facebook
Twitter
Pinterest
LinkedIn

Sign Up for Our Newsletter

Subscribe to us to always stay in touch with us and get latest news, insights, jobs and events!