Year: 2025
Type: article
Abstract: Background Rapid advances in general-purpose artificial intelligence are compressing automation timelines: around 40–50% of tasks in advanced economies are technically automatable, and a quarter of total hours could be migrated to machines before 2030. Fears of large-scale displacement from artificial-intelligence (AI) adoption have prompted calls for either mass reskilling or unconditional cash transfers. Another option, the emergence of a “supervisory economy” in which humans specialise in overseeing AI systems, remains empirically under-examined. Methods Using a balanced panel of 12 economies observed annually from 2014 to 2023, we construct a sector-weighted AI-exposure index and match it to labour-force data on unemployment, supervisory employment, public transfers, R&D, and GDP per capita. Two-way fixed-effects regressions are estimated, first linearly and then with a quadratic AI term to capture non-linearity in technology shocks. Robustness checks include lagged covariates, alternative normalisations and clustered standard errors. Results The preferred quadratic specification reveals an inverted-U relationship between aggregate AI exposure and unemployment: joblessness rises at low-to-moderate exposure but falls once adoption surpasses approximately 1.7 standard deviations above the mean. Supervisory depth—measured as the share of senior- and middle-management roles—has no significant standalone effect and does not significantly moderate AI’s impact. Higher real income consistently dampens unemployment, while public transfers and contemporaneous R&D outlays show limited short-run cushioning. Conclusions National labour markets appear to follow a concave AI trajectory: modest automation initially displaces routine labour, but at high penetration, complementary demand offsets job losses. However, simply expanding managerial layers is insufficient to buffer early shocks; effective resilience likely hinges on coupling oversight capacity with targeted technical upskilling and productivity-enhancing growth. Policy mixes that rely primarily on cash transfers or untargeted innovation spending risk delaying, rather than mitigating, employment re-equilibration in the age of AI.
Source: F1000Research
Institution: University of Canberra
Language: English
Subfield: Economics and Econometrics
Domain: Social Sciences
Open Access status: gold
Estimated APC paid: $1,350
OpenAlex ID: W4415914474