Article by Madeleine I. G. Daepp, Kiran Tomlinson, Scott Counts & Siddharth Suri: “Knowledge work has been key to economic flourishing in most advanced and many emerging economies in the last half century. Defined by the synthesis and creation of ideas rather than the production of physical goods, knowledge work involves the processing of non-routine problems that require judgment-based and creative intellectual capabilities. Such work is a large and important component of contemporary economies, accounting for an estimated third to half of all jobs in high-income countries and a fifth of all jobs globally. Achieving sustained economic growth increasingly depends on the ability to leverage and create knowledge, with countries actively seeking to transition to knowledge economies to improve their economic outcomes. Knowledge work is also the foremost application for which workers are using generative artificial intelligence (AI). A critical question for the future of twenty-first-century economies, then, is whether generative AI could democratize knowledge work by expanding the set of people who can engage in and benefit from it.
Generative AI’s effect on knowledge work hinges on emerging challenges along those two dimensions: (1) who benefits from AI’s use and (2) who actually uses AI. In this Perspective, we synthesize recent empirical work to map out these challenges and describe both technical and policy interventions to mitigate harms and ensure that benefits are widely shared. Technological and institutional fixes will need to be developed in tandem. Policies will need to be calibrated—towards either sharing productivity gains or building skills—according to what current models enable, and tooling will need to be made broadly usable if AI literacy and adoption pushes are to be effective in closing persistent participation gaps…(More)”.
