Productivity growth is determined by multiple factors, including the rate of technological progress, and few shifts have been as rapid as the rise of artificial intelligence (AI).
A recent report from Lightcast highlights the rapid transformation of the workforce by AI. Drawing on global job postings, the report reveals that demand for Generative AI (Gen AI) skills in non-tech roles has surged by 800% since the launch of ChatGPT in 2022. At the same time, the average job has seen 32% of its skill requirements change in just 3 years.
Closer to home, the Gen AI Capacity Study by Jobs and Skills Australia found that Gen AI offers major productivity potential, alongside clear challenges. Adoption is still in its early stages, with 57% of ASX200 companies referencing AI-related investments in 2024. Early signs suggest that while AI will reshape how we work, it is more likely to augment, not replace jobs (especially in high-skilled occupations where human oversight remains essential).
Both reports emphasise that we’re not being replaced by bots just yet. Of the top 10 most in-demand skills identified by Lightcast data, only two were AI-specific. The rest were human capabilities, including analytical thinking (growing in importance as AI handles routine tasks), customer service, and (perhaps somewhat surprisingly) writing (with AI amplifying human communication but not replacing it).
The National Artificial Intelligence Centre adds nuance to the story. It describes Australia’s AI innovation as evolutionary rather than revolutionary (emerging from existing industry capabilities, rather than in isolation). And it isn’t necessarily a smooth road, with research indicating that 80% of AI projects fail to progress beyond pilot (double the failure rate of conventional IT projects). Barriers include insufficient governance, immature digital infrastructure, unclear human-AI roles, and weak case selection. Additionally, 93% of businesses report struggling to measure return on investment and face risks such as output accuracy, intellectual property, cybersecurity, privacy, and compliance.
The message seems clear – AI holds promise, but productivity gains won’t come from algorithms alone. They will depend on how quickly businesses invest in skills, how effectively workers adapt, and how well policy keeps pace.