Artificial intelligence will take another three to five years to deliver broader productivity gains, despite continued hiring across the economy, according to Yahoo Finance's Market Hang, published Nov. 6.
The gains depend on businesses using the technology more widely. Current improvements in output per worker were attributed to changes made during the pandemic, before the latest generation of AI became available.
“We need to see that integrated into the real economy before we get that big payoff,” according to the discussion.
The discussion began with employment, where continued hiring was described as consistent with an expanding economy and stronger corporate profits. Wage growth and the share of people working or seeking work were also identified as measures that matter beyond a single month's hiring.
Retirements among baby boomers and restrictions on immigration were described as limits on the supply of workers. Those constraints mean fewer additional jobs are needed each month to keep the labor market stable.
Inflation was still running ahead of wage growth, reducing what workers could buy with their pay. Consumer spending nevertheless continued, leaving open how long households could sustain purchases if their purchasing power kept falling.
Asset ownership also divided households' experience of the economy. People who already owned property were described as being in a different financial position from those still trying to buy a home.
The discussion then turned to whether company job cuts showed that AI was reducing employment. A distinction emerged between savings at individual businesses and changes large enough to affect hiring across the economy.
Technology companies' earlier hiring during the pandemic complicated that assessment. Workforce reductions could reflect attempts to undo that expansion as well as changes associated with AI, making layoffs alone an uncertain measure of the technology's effects.
Productivity, the amount produced for each hour worked, offered another test. The discussion left open whether businesses could continue adding workers while also getting more output from AI, with recent productivity figures described as volatile and subject to revision.
Attention then shifted to investment. Some client money went into additional purchases during the previous day's selling, while the longer construction schedules behind AI projects remained central to questions about when spending would produce returns.
Financing costs created another division between businesses. Companies able to fund construction from cash generated by existing operations were described as better placed to absorb costs over several years than businesses dependent on borrowing.
The timing and size of AI's eventual payoff remained unresolved. So did whether investment returns would cover the costs of building the infrastructure needed to support it.
