MIT professor hoses down predictions AI will put a rocket under the economy

It's easier to foresee growing inequality than surging growth

Artificial intelligence may not do much to boost productivity – and could end up widening the income gap between owners of capital and workers.

In a US National Bureau of Economic Research paper titled "The Simple Macroeconomics of AI," Daron Acemoglu, professor of economics at Massachusetts Institute of Technology, argues that predictions AI will improve productivity and boost wages in a "blue-collar bonanza" are overly optimistic.

"AI will have implications for the macroeconomy, productivity, wages and inequality, but all of them are very hard to predict," Acemoglu argues. "This has not stopped a series of forecasts over the last year, often centering on the productivity gains that AI will trigger."

One rosy forecast predicts the advent of AI will see 100 percent GDP growth over the next decade, and a more modest forecast from Goldman Sachs of "seven percent (or almost $7 trillion) increase in global GDP and lift productivity growth by 1.5 percentage points over a ten-year period." Similar optimism was evident in last year's McKinsey Global Institute report, which suggested AI and other automation tech could increase annual average GDP growth by 0.5 to 3.4 percentage points in advanced economies over the next decade.

Acemoglu is skeptical, noting that prior introductions of automation through robotics benefited business owners and managers while workers experienced more negative outcomes.

Based on research last year that estimated about 20 percent of US workers could have half of their jobs done by an LLM, Acemoglu estimates that AI can save 27 percent in labor costs, or 14.4 percent in overall costs.

But those figures don't necessarily do much for productivity.

"This calculation implies that total factor productivity (TFP) effects within the next ten years should be no more than 0.66 percent in total – or approximately a 0.064 percent increase in TFP growth annually," Acemoglu reasons.

The professor therefore anticipates AI will boost GDP growth by only 0.93 percent to 1.16 percent over the next decade.

But even that figure may be too optimistic, he argues, because productivity estimates come from automating "easy tasks" – future tasks may be more complicated and less amenable to automation. He therefore contends there will be a more modest increase in TFP and GDP in the next ten years – on the order of 0.53 percent and 0.90 percent, respectively.

And some of that GDP growth may not improve overall economic welfare if the investment in AI brings with it extra costs, like higher energy consumption requirements.

Acemoglu goes on to argue that AI is unlikely to significantly improve wages and that even if the technology improves the productivity of low-end and middle-performing workers, it may not reduce inequality.

"I estimate that AI will not reduce inequality and is likely to have a negative effect on the real earnings of low-education women (especially white, native-born low-education women)," he asserts in his paper. His findings also suggest that "AI will further expand the gap between capital and labor income as a whole." ®

More about


Send us news

Other stories you might like