AI’s Efficiency Boom Is Still All Forward Guidance | American Enterprise Institute

The ultimate payoff from the 2020s AI boom will look a lot like the payoff from the 1990s internet boom: companies using the technology to become more productive.

Well, if it happens. 

All that extraordinary spending on chips and data centers, while wonderfully profitable for chip-makers and cloud providers, is merely the opening act. Granted, it’s been an incredible opening act that isn’t ready to leave the stage. In a new note, the bank Goldman Sachs sees AI investment of nearly $600 billion in the US next year, close to 2 percent of GDP.

At some point, however, the infrastructure spending phase has to give way to full-scale deployment. This is when firms far from Silicon Valley find practical, profitable ways to use AI in a big way.. 

A new analysis from the St. Louis Fed is the latest to suggest that we aren’t yet at that inflection point, at least based on what businesses are saying—which isn’t to say companies are ignoring AI. They talk about it a lot, notably in corporate earnings calls, which is how the economists who wrote the analysis address the issue. The team scanned some 490,000 transcripts from 5,198 publicly traded US firms running from 2000 through 2025.

Before ChatGPT, businesses pretty much never talked about AI when they talked about productivity. By late last year, however, about 15 percent of productivity-related sentences also mentioned AI, a number that seems likely to be higher today. 

Moreover, while businesses used to talk about “generative AI,” now more and more they’re referencing “using AI” and “AI tools,” suggesting a greater focus on the productive adoption of the technology. Indeed, the phrase “generative AI” has fallen out of fashion. I would speculate its use today marks a company as one not making a serious attempt to incorporate the technology, maybe something akin to using “high tech.”

Given the importance of productive adoption and return on investment to the whole Age of AI story, this to me is money finding (bold by me): 

“The majority of talk about productivity is forward-looking, especially when AI is involved. Approximately 95% of the productivity-related sentences in earnings call transcripts that mention AI refer to future gains. … For now, earnings calls point less to broad realized productivity gains than to a corporate sector actively investing in, experimenting with and expecting future gains from AI. We will continue to monitor whether AI-related productivity discussion in upcoming earnings calls shifts from future-tense expectations to realized gains.”

So still more of a Tomorrowland thing than a Todayland. Anyway, there’s your assignment, AI optimists. Don’t wait for government stats to show AI-driven productivity gains. Check out those earnings calls.