Automation and the changing nature of work

Area of Science:

• Economics
• Labor Market Analysis
• Artificial Intelligence Impact

Background:

• Automation poses a significant challenge to the future of work, necessitating an understanding of job vulnerability.
• Existing classifications of automatability require empirical validation through data-driven methods.
• Identifying specific job attributes that predict automation risk is crucial for workforce planning.

Purpose of the Study:

• To identify key job attributes, particularly skills and abilities, that predict a job's likelihood of being automated.
• To analyze the relationship between specific cognitive, social, and physical skills and automation risk.
• To leverage a machine learning approach for a nuanced understanding of automation susceptibility.

Main Methods:

• Utilized the Josten and Lordan (2020) classification of job automatability.
• Employed European Labour Force Survey data for comprehensive labor market insights.
• Applied a machine learning regression model to predict automation likelihood based on job attributes.

Main Results:

• Skills related to non-linear abstract thinking were found to be most resistant to automation.
• Jobs involving 'people' engagement combined with cognitive ('brains') tasks showed lower automation risk, often requiring soft skills.
• Highly physical jobs, particularly those involving object manipulation, are most susceptible to automation, unless they incorporate cognitive or interpersonal elements.

Conclusions:

• The future of work is shaped by the interplay of cognitive, social, and physical demands within jobs.
• Cultivating abstract thinking and soft skills is paramount for enhancing job security in an automating world.
• Policy and training initiatives should focus on skills that complement, rather than compete with, automation technologies.