Artificial intelligence didn’t suddenly emerge in November 2022 when OpenAI rolled out ChatGPT. The concept of AI and its foundational technologies have roots dating back to the 1950s. Indeed, computer advances that were not thought of as “AI” decades ago might now be seen as crucial to the technology’s development. Expert systems, early machine learning algorithms, and other computer-based decision-making tools were the precursors of modern AI, including deep learning and neural networks.
By employing a broad and historically sensitive definition of AI — one encompassing earlier forms of machine intelligence and automation — economists at the Bank of Italy traced AI-related innovation over several decades. Their approach allowed them to analyze patents filed from 1980 through 2019, retroactively classifying them based on AI content. This comprehensive view enabled the researchers to study the long-term macroeconomic impacts of AI development, from its early stages to recent advancements. Their goal: Understand how AI has gradually reshaped the economy, labor markets, and productivity over time, rather than focusing solely on recent breakthroughs.
What the researchers found paints a surprisingly rosy picture — well, surprising a society so steeped in techno-pessimism about “robots taking all the jobs” — of AI’s macroeconomic impact. As AI innovation surges, it ripples through the economy, boosting industrial output and driving down consumer prices. Far from the job-killing specter often portrayed in popular media, AI appears to be a job creator, pushing up employment rates, extending work hours, and fattening paychecks.
Perhaps most critically, it supercharges overall economic efficiency, as measured by total factor productivity. These benefits aren’t confined to Silicon Valley or high-tech hubs; they permeate diverse sectors of the economy, suggesting AI’s positive influence extends far beyond the realms of chatbots and autonomous vehicles. This is great news because higher productivity will translate into higher living standards and wages over the long run, if history is any guide.
True, the study concludes, “the expansionary impact of AI-driven technological development comes at the cost of an increase in inequality across income and wealth distributions.” But that doesn’t mean all boats won’t rise even if some rise faster than others. I would also note that early studies of generative AI show that it most helps the productivity of lower-skilled workers, which might counteract some of the inequality impacts described in the Bank of Italy paper.
The post Think about the Economic Impact of AI … appeared first on American Enterprise Institute – AEI.