Economists are waiting for data to indicate AI has delivered on its promises of productivity gains, but that moment is likely years away, according to Jim Reid, Deutsche Bank Research Institute global head of macro and thematic research.

Reid predicted that AI would, indeed, create new jobs and increase workplace efficiency, but said people are being a little overambitious in their timelines of when the technology will have a rippling impact on the economy.

“In my career I haven’t seen anything like AI in terms of potential for productivity,” Reid told Bloomberg Television on Tuesday. “But I would probably caution that it is going to take a number of years for us to properly embed it into enterprises to really get the benefits of that.”

Multiple economic indicators suggest that despite tech leaders touting AI’s ability to overhaul the workforce, there’s little evidence of its widespread impact so far. As of last month, the Yale Budget Lab noted no significant changes in occupational mix or length of unemployment for roles with high AI exposure, leading researchers to conclude no evidence of AI-related labor market disruptions. Apollo chief economist Torsten Slok noted in a recent blog post, citing Bloomberg and Macrobond data, that while profit margins for the Magnificent Seven increased from about 15% to 25% between the first quarters of 2023 and 2026, margins for the rest of the S&P 500 index were around 10% over the same period. The data suggests that while tech companies are able to more easily integrate new technology into their operations, other industries have been slow to deploy the technology, let alone see its benefits.