Saving your brains for the good stuff, not ‘grunt work’

Disclosure Statement: Durand Financial Services Pty Ltd and its advisers are authorised representatives of Fortnum Private Wealth Ltd ABN 54 139 889 535 AFSL 357306. General Advice Warning: The information contained within this website does not consider your personal circumstances and is of a general nature only. You should not act on it without first obtaining professional financial advice specific to your circumstances.

Electricity networks and banks will need high-tech tools to overhaul decades of legacy systems that could otherwise leave the nation vulnerable to blackouts and hackers.

Developer platform GitHub prides itself on being home to more than 100 million open-source developers including coders in Australia.

Businesses are increasingly adopting a code completion tool called Copilot developed by GitHub and American artificial intelligence organisation OpenAI.

And yet almost half of Australian workers (47 per cent) say they have never used generative artificial intelligence (GenAI) in their role, and almost three-quarters (73 per cent) say they do not believe it is relevant to their role.

This is despite the Australian Computer Society predicting GenAI will impact most (86 per cent) of occupations.

“From a consumer perspective, everyone has had some exposure to these types of tools – to ChatGPT, Siri or Alexa – but they’re not really sure how to get started,” GitHub chief operating officer Kyle Daigle told AAP during a visit to Australia.

“Many businesses are hearing the demand from their developers … they’re completing their tasks up to 42 per cent faster because Copilot helps them complete their thought.

“In the energy and manufacturing sectors, every company at some point is a software company.”

He said new workers needed to be quickly brought up to speed, and AI could bridge the gap for engineers who might not have a background in dealing with energy systems or overhauling the back end of a bank.

Copilot can act as a mind-reader to help unravel old code written for banks or hospitals in languages that most of today’s coders no longer use.

“No one’s going to say ‘hey I would love to migrate five million lines of (65-year-old computer code) Cobol to (more recent programming language) Java’,” Mr Daigle said.

Power networks and banking systems in particular are pushing through the fear factor to write millions of lines of code for everyday transactions and activity.

“Take advantage of your Copilot and Copilot Studio and the innovations that mean you can have less bodies doing the grunt work and more brains doing the good work,” Horizon Power boss Stephanie Unwin recently told Australian Energy Week.

“You’ve got to have an eye forward to things like AI … there’s definitely real opportunities to use labour in a different way,” she said.

Commonwealth Bank was among the first in Australia to test Copilot to help deliver code faster, protect the bank, and get more done for its customers.

But keeping a human in the loop, rather than handing decision-making to a machine, remains important.

Three-quarters of the bank’s software engineers rated the GenAI tool “very helpful” and accepted nearly 80,000 lines of code that Copilot had recommended – one-third of all the recommendations made.

Copilot has also steered software engineers at ANZ Bank toward improved productivity and code quality, providing detailed suggestions while keeping decisions with the developers.

“Given the unique circumstances of the workforce and economy here in Australia, there’s more of an interest in giving it a try,” Mr Daigle said.

He said Australia was interesting because banks were not wasting time on a big plan or scaring themselves with a big rollout.

“That does give the Australian market a leg up, both for the services you’re providing inside your country and all the SaaS (service as a software) companies that serve the world from Australia,” he said.

 

Marion Rae
(Australian Associated Press)

0

Like This