Stay informed with free updates
Simply sign up to the Artificial intelligence myFT Digest — delivered directly to your inbox.
The writer is director of Stanford University’s Digital Economy Lab and co-founder of Workhelix
For over a decade, economists have grappled with a modern iteration of the Solow Paradox: we have seen artificial intelligence everywhere except in the productivity statistics. Sceptics argue that the reason for this is that modern innovation in machine learning systems and now generative AI pale in comparison to the great inventions of the past. However, the latest benchmark revisions from the Bureau of Labor Statistics suggest the statistical fog may finally be lifting.
Data released this week offers a striking corrective to the narrative that AI has yet to have an impact on the US economy as a whole. While initial reports suggested a year of steady labour expansion in the US, the new figures reveal that total payroll growth was revised downward by approximately 403,000 jobs. Crucially, this downward revision occurred while real GDP remained robust, including a 3.7 per cent growth rate in the fourth quarter. This decoupling — maintaining high output with significantly lower labour input — is the hallmark of productivity growth.
My own updated analysis suggests a US productivity increase of roughly 2.7 per cent for 2025. This is a near doubling from the sluggish 1.4 per cent annual average that characterised the past decade.
This shift aligns with the productivity “J-curve” that my colleagues and I have explored in earlier research. General-purpose technologies, from the steam engine to the computer, do not deliver immediate gains. Instead, they require a period of massive, often unmeasured investment in intangible capital — reorganising business processes, retraining the workforce and developing new business models. During this phase, measured productivity is suppressed as resources are diverted to investments. The updated 2025 US data suggests we are now transitioning out of this investment phase into a harvest phase where those earlier efforts begin to manifest as measurable output.
Micro-level evidence further supports this structural shift. In our work on the employment effects of AI last year, Bharat Chandar, Ruyu Chen and I identified a cooling in entry-level hiring within AI-exposed sectors, where recruitment for junior roles declined by roughly 16 per cent while those who used AI to augment skills saw growing employment. This suggests companies are beginning to use AI for some codified, entry-level tasks.
While the trends are suggestive, a degree of caution is warranted. Productivity metrics are famously volatile, and it will take several more periods of sustained growth to confirm a new long-term trend. Furthermore, powerful macroeconomic headwinds, ranging from geopolitical trade wars to fiscal or monetary mismanagement, could counteract these efficiency gains.
But there is cause for further optimism when we distinguish between potential and realised gains. Many businesses are using generative AI for only a small fraction of tasks. Some merely employ AI for translation or summarisation — what might be called “glorified dictionary” use.
Conversely, my company found a small cohort of power users are leveraging interactive conversations with AI agents to automate end-to-end workstreams, such as generating complete marketing plans, compressing weeks of effort into hours. The challenge for businesses is not simply acquiring the technology but using it to level up the average employee. That will boost not only their own profits but productivity gains across the economy.
We are transitioning from an era of AI experimentation to one of structural utility. We must now focus on understanding its precise mechanics. The productivity revival is not just an indicator of the power of AI. It is a wake-up call to focus on the coming economic transformation.