AI’s Productivity Paradox: How the 1980s Solow Effect Predicts a Coming Economic Tipping Point

There’s an old joke among economists: “You can see the computer age everywhere but in the productivity statistics.” It wasn’t funny in 1987 when labor economist Robert Solow wrote those words. Yet they were true. Personal computers, corporate mainframes, and the first vestiges of the modern internet dominated headlines, yet productivity stagnated. This became known as Solow’s Paradox.

By the mid-1990s, the paradox resolved itself. Productivity surged, tech wealth exploded, and—despite crashes and recoveries—technology became the backbone of the modern economy. Today, AI appears to be following the same trajectory. And new data suggests a similarly massive productivity—and wealth—tipping point may be imminent.

AI’s Solow Paradox: Early Hype, Limited Impact

Since generative AI entered mainstream use with ChatGPT in 2022, it has mirrored early computing trends. The world buzzes about large language models (LLMs) and artificial general intelligence (AGI), yet the economic impact remains underwhelming.

Even the most prominent AI companies generate modest revenue. OpenAI, for example, reported annualized revenue of around $20 billion by the end of 2023. For context, that’s roughly the size of the pest control industry—and half the size of the pizza industry.

The gap between excitement and economic reality is evident in broader data. A February study surveyed 6,000 business leaders on AI’s impact. Despite 63% reporting AI adoption, 90% saw no effect on employment or productivity.

Official statistics echo this disconnect. Research from the Federal Reserve Bank of St. Louis found generative AI delivered only a 5.4% productivity improvement—far below the transformative gains promised by AI valuations.

Solow’s Paradox, it seems, has returned.

Signs of a Turning Point: AI’s Economic Breakthrough

New data indicates this may be changing. While still early, recent earnings reports and studies suggest AI is beginning to deliver measurable economic value.

Alphabet (Google’s parent company) provided the strongest evidence in its Q1 earnings. The company reported that AI boosted core Search revenue by 19% and Google Cloud revenue by 63%. Even more telling, Alphabet stated that AI enterprise technology drove the majority of Cloud’s gains, with AI-driven revenue from major clients up 800% year-over-year.

Microsoft also reported significant AI-driven revenue in its latest earnings. The company’s AI business now generates an annual run rate of $37 billion, largely fueled by enterprise adoption.

Smaller AI players like Salesforce, ServiceNow, and Databricks have similarly reported AI-related revenue growth, signaling a broader shift toward tangible economic impact.