A Task- and Attribute-Based Approach to Understanding the Effects of Artificial Intelligence on Worker Displacement and Training Strategies

Rick Miller

doi:https://doi.org/10.21985/n2-sxmv-wd10

Work

A Task- and Attribute-Based Approach to Understanding the Effects of Artificial Intelligence on Worker Displacement and Training Strategies

Public Deposited

Download PDF

Download All Files (.zip)

Advances in machine learning and artificial intelligence (AI) are enabling computers and robotics to take on an increasing number of work-related tasks that previously were the sole domain of humans. This trend raises questions: Which jobs will be most susceptible to replacement by automation? How many workers risk being replaced? What traits are exhibited by workers poised for success in an economy where AI and robotics play a primary role in productivity? This paper explores findings from six academic papers and two large-scale primary research studies conducted by for-profit and not-for-profit knowledge centers to understand the quantitative effect AI will have on the labor force; it explores findings from two book-length works and two academic papers on the traits and attributes human employees will need so they can work alongside AI-based systems successfully, and it leverages several predictive modeling techniques (linear regression, logistic regression, linear discriminant analysis, quadratic discriminant analysis, and random forest) to identify workers who are poised to succeed alongside automation.

Last modified