While AI companies continue to command the global economy and pour billions of dollars into building data centers, it’s still not even clear that their products can deliver on one of their core premises: increasing productivity.
One big skeptic of the tech’s economic promises is vice president and principal analyst at the global market analytics firm JP Gownder, who says we’re simply “not seeing” an AI-driven boost to productivity in any of the available data.
“You begin to get the picture that information technology isn’t measured always in as linear a way into productivity as people assume,” he told The Register in a new interview. “It just isn’t there.”
This isn’t a statement on how transformative AI might or might not be. It could have massive effects without necessarily translating to economic gains, a phenomenon known as the Solow Paradox, named after the Nobel Prize winning economist Robert Solow who correctly predicted “that by 1987 the effects of the PC revolution can be seen everywhere, except in the productivity statistics,” in Gownder’s analysis.
He pointed to data from the US Bureau of Labor Statistics showing how between 1947 to 1973 — before the advent of PCS — productivity improved by 2.7 percent annually, but only 2.1 percent between 1990 and 2001, once PCs had hit the mainstream.
“So despite all those PCs, it was a lot lower,” Gownder said. “And [from] 2007 to 2019 it was 1.5 percent.”
Mountains of research as well as cases of workplace deployment of AI have suggested that the tech is far from being ready for primetime. One notable MIT study found that 95 percent of companies that integrated AI saw zero meaningful growth in revenue. For coding tasks, one of AI’s most widely hyped applications, another study showed that programmers who used AI coding tools actually became slower at their jobs.
AIs designed to automate entire tasks aren’t looking too hot either: when researchers at the Center for AI Safety tested AI agents’ ability to complete remote work tasks, not a single model was able to complete more than three percent of their assignments. The very introduction of AI into the workplace appears to be disastrous for employee relations, suggests another study, which found that AI resulted in employees passing off low quality “workslop” with the expectation that someone else down the line would polish the AI’s shoddy output.
“A lot of generative AI stuff isn’t really working,” Gownder told The Register. “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.]”
“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,” he added.
That said, Forrester’s research does predict that AI and other automation tech, like physical robots, will see a hefty six percent of jobs replaced by 2030, amounting to some 10.4 million roles.
“These jobs are lost structurally, like they’re gone for good, because they’ve been replaced,” Gownder told The Register. “That’s not an insignificant hit to the economy.”
There may come a point where employers realize that AI isn’t working out, Gownder said. Some bosses who fired workers in favor of AI agents have already eaten crow and rehired their human grunts. But “AI” may simply be a way of papering over other forms of cost-cutting.
“Outsourcing is a very popular one,” he told The Register. “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.”
More on AI: Researchers Just Found Something That Could Shake the AI Industry to Its Core