Is GenAI’s Affect on Productivity Overblown?

We don’t know great about how LLMs trade productivity at the company stage — and a end reading of most contemporary stories suggests they’ll hinder as great as assist.

January 08, 2024

Illustration by Kemal Sanli

Generative AI worship LLMs were touted as a boon to collective productivity. Nonetheless the authors argue that leaning into the hype too great would possibly well moreover very effectively be a mistake. Assessments of productivity in total care for the duty stage and the design in which other folks would possibly well moreover exhaust and like LLMs. The exhaust of such findings to scheme gargantuan conclusions about company-stage performance would possibly well moreover yell costly. The authors argue that leaders wish to worship two core concerns of LLMs before adopting them firm-wide: 1) their continual ability to carry out convincing falsities and 2) the likely long-length of time destructive effects of using LLMs on workers and inner processes. The authors outline a long-length of time point of view on LLMs, to boot to what styles of duties LLMs can place reliably.

Large language fashions (LLMs) were heralded as a boon to collective productivity. McKinsey boldly proclaimed that LLMs and a kind of styles of generative AI would possibly well moreover develop corporate profits globally by $4.4 trillion yearly, and Nielsen trumpeted a 66% amplify in employee productivity throughout the utilization of these same tools. Projections worship these like made discovering ways to make exhaust of these tools — and turbocharge productivity — a high precedence for quite loads of companies right throughout the final year. While we’re intrigued and impressed by this contemporary technology, we expose cautious experimentation other than wholesale firm-wide adoption.

Nathanael J. Immediate is the Jorge Paulo and Susanna Lemann Chair in Entrepreneurship and affiliate professor of management at the Marshall Faculty of Enterprise at the University of Southern California. He directs the USC Neely Center for Moral Management and Willpower Making and co-directs the Psychology of Technology Institute. His analysis examines guilty management within the age of AI.