AI Impacts: Learning from Recent History

When it comes to technology discussions, we are all rightly consumed with the expected impacts of AI on our jobs, businesses, and communities.  While it may be too soon to clearly delineate future impacts, researchers are beginning to offer some useful insights.   A recent National Bureau of Economic Research (NBER) paper by some of America’s leading AI researchers offers some very interesting perspectives from recent US history.  The study taps into the 2018 US Census Survey of Business, which surveyed more than 850,000 US businesses on their use of technology.  The researchers examined how firms were using AI-related technologies such as machine learning, voice recognition, and natural language processing.  Because the survey dates to 2018, this data tracks “early adopters” of these new technologies.

The study found that only 6% of US firms had adopted AI technologies at that time.  However, most large US firms were actively using these tools, and most sectors of the US economy were being affected by AI-related technologies. Not surprisingly, the study found that AI use was closely tied to high-growth entrepreneurship, and that the leaders of these firms tended to be (not surprisingly) younger, better educated, and more tech-savvy.  AI adoption was also heavily clustered in a few tech hot spots and “superstar” cities.  The authors conclude that these patterns could presage an “AI Divide,” where some communities and regions may lack full access to the benefits of AI technology.