You can also access this resourceand many others, in Issue 370 of my Talent Edge Weekly newsletter
Much has been shared about what AI can automate or augment and its impact on jobs and skills. A new working paper by The Burning Glass Institute economist Soumitra Shukla examines the other side: tasks AI cannot yet perform or sufficiently accelerate. Shukla notes that as AI speeds up some tasks, others that still require significant human effort can become bottlenecks, limiting overall productivity gains. Using 288 million U.S. job postings, the paper identifies which occupations may face the greatest productivity constraints from tasks that AI cannot yet perform or sufficiently accelerate. Consider my recruiting example: AI may accelerate sourcing, initial screening, and scheduling. Yet human-dependent work, such as assessing candidates against business needs and making hiring decisions, can limit gains in hiring quality and time to fill. In other words, AI can make individual tasks significantly faster without necessarily improving the speed or effectiveness of the end-to-end process. Five questions HR teams can use to help their organizations identify and address potential bottlenecks: 1) Which tasks is AI making faster? 2) Which remaining tasks aren’t keeping pace? 3) How are these slower tasks affecting the overall process and its outcomes? 4) What capabilities, processes, or organizational changes could address the bottlenecks? 5) How will we know whether these changes improve end-to-end performance? To make this manageable, start with one process where your organization sees the greatest opportunity for AI to improve outcomes. Use these questions to identify potential bottlenecks, guide decisions on where to focus attention, and assess whether expected benefits materialize.
Source: www.brianheger.com
