Early AI adopters face a brutal truth: productivity tanks before it soars. A new study of 30,000 U.S. manufacturers reveals this counterintuitive pattern.
Companies that embraced AI between 2017 and 2021 first watched their productivity plummet. The culprit? AI disrupted their finely-tuned operations, like just-in-time inventory systems.
But those who survived the initial chaos emerged stronger. These companies eventually outperformed their peers in sales, productivity, and even hiring. The catch? Many firms didn't make it through the turbulent transition.
Older, larger companies struggled the most. "Surviving this seems like part of the problem," says Kristina McElheran, the University of Toronto researcher behind the study.
The findings challenge the rosy narrative that AI simply "augments" jobs. During the study period, AI adoption crept up from 7.5% to 9.1% among surveyed firms.
ECB President Christine Lagarde added perspective: up to 29% of European workers face high AI exposure. But she expects job destruction to balance with job creation.
Why this matters:
The AI productivity paradox mirrors the 1990s computer revolution - new tech often hurts before it helps
Companies rushing to adopt AI might want to pack a parachute - and some patience
