The “smart money” has weighed in on potential AI employment impacts. The best way to describe the new estimates from Goldman Sachs Research (GSR) is that they reflect the nature of how technological change actually happens, minus the techno-optimism or doomerism.
GSR estimates that, as AI adoption accelerates, US unemployment could rise by about 0.5 percent. With a workforce of 168 million, that would translate into 860,000 newly unemployed people, a figure that is neither insignificant nor catastrophic. It’s also consistent with what we typically see when new labor-saving technology enters the workplace: a short-term dip followed by higher employment as aggregate economic demand, driven by higher productivity and income.
Time is on the side of current workers. GSR says if we could somehow apply AI at its current capability level across the entire economy all at once, 2.5 percent of workers might be displaced. That would be the equivalent to a serious recession throwing 4.2 million out of work. The good news is that such an instantaneous rollout is impossible. AI isn’t frosting on a cake; it’s a complete revision of the recipe requiring new business processes and substantial investments in tangible as well as human capital to make it work, both requiring time and money. At the outer edge of possibility, broad adoption could affect six to seven percent of the workforce. That sounds dramatic, but Goldman’s economists stress that even these disruptions are likely to be temporary, as displaced workers eventually find new opportunities in sectors AI helps create or expand.
Still, in the short and medium term, some occupations are more AI-exposed than others. Computer programmers, accountants, legal assistants, and customer service representatives—roles where routine, code-able tasks predominate—will feel the heat of change first. Programmers have been on the leading edge of this trend and, by anecdote and some early data, are having a harder time finding work. GSR notes that young tech workers (those between 20 and 30) are already experiencing hiring slowdowns, with unemployment rising nearly three points this year.
AI’s larger story will be about productivity. Generative AI is projected to raise labor productivity in advanced economies by roughly 15 percent once fully integrated into daily work, improving incomes and living standards as well as laying the foundation for future generations to reap the benefits of growth. Nonetheless, some will face immediate job losses. The key to smoothing the transition is to pay attention to those who are impacted and, through policy and programs, make clear workers will not be left behind as we make the leap to an AI-infused economy.
