AI creates jobs, AIB reports

AI is creating more jobs than it is taking, according to an Australian Institute of Business (AIB) report.

Net availability of highly automatable AI-exposed jobs grew by 10 per cent 2019-2024. (Credit: Australian Institute of Business)

Net AI-exposed job availability grew by 10 per cent over the 2019-2024 analysis period.

Roles that are exposed to AI are those where tasks may be augmented or automated, such as ICT professionals, clerks, call centre operators or even teachers and lawyers. 

AI-competent workers earned $39,000 more on average than their peers, according to the AIB report.

Jobs that could be automated by AI grew by 45 per cent and augmentable roles – where a worker is supplemented by AI – grew by almost half.

A government report from Job and Skills Australia last year found that generative AI was more likely to augment jobs than replace them.

That report stated “limited evidence” for AI-related layoffs in Australia, but acknowledged some firms had reported “technological change” as the reason for group layoffs.

It noted some voice acting work and content recording, production and editing roles were being replaced by generative AI.

An Australian Association of Voice Actors’ submission to the Senate Inquiry on Adopting Artificial Intelligence relayed a voice over talent agency’s concerns that “an entire piece of content can now be developed using AI”.

The AIB report supports CSIRO research, which found that non-AI job postings for AI-adopting firms grew 36 per cent faster than for non-adopting firms between 2020 and 2023.

“Achieving productivity gains isn't just about doing things faster; the big gains come from doing things in a new way or finding new ways of adding value using AI,” Dr Claire Mason, leader of the CSIRO work and productivity team, said.

She said the excitement about what is theoretically possible with AI did not directly translate to jobs being replaced.

“Planes can fly themselves and have been able to for a really long time. And yet, we still have two human pilots in every plane,” she said.

Dr Mason said research had shown that demand for workers in AI-adopting firms remained strong, but had weakened for some AI-exposed workers in non-adopting firms.

Reduced demand for non-AI-adopting workers could be because they are potentially less productive than their AI-adopting peers, Mason said.

“The future workforce will need more than technical AI skills; professionals will also need analytical judgement, business understanding and the ability to apply AI responsibly to real organisational problems,” Dr Sajjad Shokouhyar, Discipline Leader in Operations and Supply Chain Management at AIB, said.

Relevant skills include communication and metacognition, which means understanding AI’s strengths and flaws so it can be applied to appropriate tasks, Mason said.

She said responsibility and diligence were also important attributes, so that AI-adopting workers would verify the quality of AI output before using it.

“New technologies usually create more new jobs than they replace,” she said.

Dr Shokouhyar said that task-relevant expertise, such as familiarity with industry-specific terminology, was an important skill allowing professionals to prompt better output from generative AI.