The Future of Employment Revisited: How Model Selection Determines Automation Forecasts

The uniqueness of human labour is at question in times smart technologies. 250 years-old discussion on technological unemployment reawakens. Prominently, Frey and Osborne (2017) estimated that half US employment will be automated by algorithms within the next 20 years. Other follow-up studies conclude only a small fraction workers replaced digital main contribution our work to show diversity previous findings regarding degree job automation is, large extent, driven model selection not controlling for personal characteristics or tasks. For case study, we consult experts machine learning industry professionals susceptibility technologies Austrian market. Our results indicate that, while clerical computer-based routine jobs are likely change decade, professional activities, such as processing complex information, less prone change.