This pilot study presents an innovative artificial intelligence (AI) model to predict lawyers' appraisal ratings in a law firm. Methodology development was based on an 11-years database comprising multiple descriptors from 229 lawyers. The AI model builds upon law firms' tournament, simulating lawyers' career competition to predict performance rankings. Within a one-year lag, the accuracy of the model was approximately 88%. With two- and three-year lag times, the predictions show only a minor drop in performance. Benefits of this in-silico strategy involve decreasing the frequency of appraisals linked with considerable time and resource savings. By highlighting the most relevant performance predictors in the firm, practitioners may identify bias in appraisals and realign talent management with business strategy. This longitudinal study aims to pilot predictive research for AI models in talent management in law firms. Future research may lead to predictive models supporting talent strategies and practices.
Abstract The long-term changes in the aggregate tasks (activities) of employees in the German automotive industry are analysed. The central question is whether the observed changes in activities occur within occupations or are driven by the reallocation of employees between occupations. International studies show that most of the aggregate changes occur within occupations. Our findings confirm these results for Germany. They also show that many occupations have the capacity to adapt and that the fear of joblessness due to technological change seems disproportionate.