Goldman warns AI productivity payoff may lag until 2030
Investors pricing an imminent AI-driven productivity boom into equities are likely miscalibrated on timing, with Goldman Sachs research suggesting meaningful macro gains will not materialize until the end of the decade due to lagging organizational investment and widespread worker resistance.
Goldman Sachs economist Elsie Peng warns that investors betting on an AI revolution are badly miscalibrated on the timeline for macroeconomic productivity gains. Comparing the current cycle to the personal computer era, she notes it took 15 years after the PC's 1981 commercialization for a broad-based productivity boom to appear in the data. If ChatGPT's 2022 launch marks the equivalent starting point, the actual economic payoff will not arrive until around 2030 at the earliest.
Peng's analysis reveals a historical J-curve: a modest productivity drag for the first four years of a new technology, flatlining for another four, and statistically significant gains only emerging in year eight. During the 1980s, surging investment in information and communications technology yielded no immediate productivity improvement. The peak impact of roughly 0.6 percentage points did not materialize until year 12.
The primary bottleneck was rarely the hardware itself. Goldman estimates each dollar of ICT hardware required at least $1.70 of complementary investment in software, data systems, and organizational overhaul. While AI hardware investment is currently outpacing the 1980s ICT buildout, spending on reorganizing work processes is lagging behind historical equivalents by a wider margin.
This organizational deficit is being compounded by active workforce resistance. An April survey of 2,400 knowledge workers found 29% admit to actively sabotaging their company's AI strategy, a figure that jumps to 44% among Gen Z employees. A separate 14-country survey found more than 54% of workers bypassed corporate AI tools in the past 30 days to complete tasks manually.
Harvard Business School researchers have termed this dynamic "symbolic adoption," where employees comply on the surface while quietly undermining the technology. The motive is largely self-preservation: 30% of self-described saboteurs fear losing their jobs to AI, and 69% of executives in the same survey confirmed their companies are already conducting AI-related layoffs. Employment for workers aged 22 to 25 in AI-exposed occupations is already shrinking more than 4% annually, according to tracking by Stanford's Erik Brynjolfsson and ADP Research.
The financial stakes of this delay are immense. Apollo Global Management Chief Economist Torsten Slok argues that AI "has been the one thing holding up both the economy and markets." He warns that a slower payoff "would risk tipping the economy into recession and the S&P 500 into a correction." While Goldman maintains that AI will meaningfully boost growth over the next decade, the path to those gains is proving much messier than current equity valuations suggest.
