Despite the billions poured into AI infrastructure, a prominent industry analyst is casting doubt on whether the technology actually delivers the productivity gains it promises. JP Gownder, vice president and principal analyst at Forrester, argues that current data simply does not support the narrative that AI is boosting economic efficiency.
In a recent interview with The Register, Gownder pointed to U.S. Bureau of Labor Statistics figures showing that productivity growth actually slowed after the widespread adoption of personal computers. From 1947 to 1973, productivity rose at an annual rate of 2.7 percent, but between 1990 and 2001—when PCs became mainstream—the rate fell to 2.1 percent. From 2007 to 2019, it dropped further to 1.5 percent.
“You begin to get the picture that information technology isn’t measured always in as linear a way into productivity as people assume,” Gownder said. “It just isn’t there.”
This phenomenon is known as the Solow Paradox, named after Nobel laureate Robert Solow, who famously observed in 1987 that the PC revolution was visible everywhere except in productivity statistics. Gownder suggests AI may be following a similar trajectory.
Evidence of AI's Limited Impact
Research appears to back Gownder’s skepticism. A notable MIT study found that 95 percent of companies that integrated AI saw no meaningful revenue growth. In coding, a widely touted AI application, another study revealed that programmers using AI tools actually became slower. Even AI agents designed to automate entire tasks struggled: researchers at the Center for AI Safety found that no model could complete more than three percent of remote work assignments.
Workplace dynamics also show friction. One study indicated that AI adoption led to employees passing off low-quality “workslop,” expecting others to fix AI-generated output. Gownder summarized the situation: “A lot of generative AI stuff isn’t really working. And I’m not just talking about your consumer experience, which has its own gaps, but the MIT study that suggested that 95 percent of all generative AI projects are not yielding a tangible [profit and loss] benefit. So no actual [return on investment.]”
He added, “It is just further context that says we’re not at a place where lots and lots of people are losing their jobs right now.”
Job Displacement Still on the Horizon
Despite the lack of immediate productivity gains, Forrester’s research predicts that AI and automation could replace six percent of jobs by 2030, equating to about 10.4 million roles. Gownder described these losses as structural, noting, “These jobs are lost structurally, like they’re gone for good, because they’ve been replaced. That’s not an insignificant hit to the economy.”
Some employers who replaced workers with AI have already reversed course and rehired humans. Gownder also suggested that “AI” is sometimes used as a cover for other cost-cutting measures, such as outsourcing. “They’re firing people because of AI, and then three weeks later they hire a team in India because the labor is so much cheaper,” he said.
The debate over AI’s economic impact is far from settled, but Gownder’s analysis serves as a counterweight to the prevailing hype, urging a closer look at the data before declaring AI a productivity panacea.
