AI Fails to Boost Productivity According to Recent CEO Study

A recent study reveals that AI's integration into the workplace has not resulted in expected productivity gains or significant employment changes, paralleling the 'productivity paradox' identified during the IT era by economist Robert Solow. Despite AI's prevalence, reflected by its mention in 374 earnings calls among S&P 500 companies, a study involving 6,000 senior executives from the U.S., U.K., Germany, and Australia found that nearly 90% reported no productivity or employment impact from AI over the past three years.

The research highlights a disconnect between the anticipated and actual outcomes of AI on productivity, reminiscent of past technological advancements. Specifically, while two-thirds of executives use AI, this only accounted for around 1.5 hours of weekly workplace use, and 25% of respondents did not use AI at all. Despite these findings, executives maintain substantial expectations for AI, forecasting a 1.4% increase in productivity and 0.8% uplift in output over the next three years.

Some researchers remain hopeful regarding AI's potential, with the Federal Reserve Bank of St. Louis reporting a 1.9% increase in productivity growth since the introduction of ChatGPT in 2022. However, the broader economic indicators, such as employment and inflation data, fail to show significant AI-driven changes, raising questions about when or if the expected returns on AI investments, which reached over $250 billion in 2024, will materialize.

Economists are drawing parallels with the IT age when the advent of new tech, such as transistors and microprocessors, also failed to boost productivity immediately, instead generating overwhelming amounts of information without effective means of leveraging it for productivity increases.