Sasha Rosenbaum

It’s been a couple of years since Large Language Models became a household name, and every company on the planet is expecting to see significant productivity gains across multiple departments from using genAI-powered tools. Yet, emerging research tells us that naively utilizing GenAI can hurt organizational productivity as much as it helps it.

This shouldn’t come as a surprise, as we have seen similar results with companies naively implementing novel technologies and methodologies before. We know how to do these things well, yet we keep falling into the trap of assuming that technology alone solves the problem, failing to account for the importance of jointly optimizing sociotechnical systems. We also tend to believe that novel systems require novel governance and processes, causing us to solve and re-solve the same problems over and over again.

In this talk, we will cover some of the common causes of productivity loss via naive adoption of genAI tools and how to take human factors into account to intelligently design sociotechnical systems to maximize genAI-powered productivity gains.