🔍 Read the full analysis: Productivity With AI: A Practical Starting Point on ThorstenMeyerAI.com
Get the latest gadgets delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
TL;DR
AI tools have produced measurable productivity gains in some workplace studies, including a reported 15% increase for customer-support agents, but that does not yet translate into a large economywide effect. The source says executives estimate AI has added about 0.2 percentage points a year to productivity growth so far; the estimate, methods and longer-term effects remain uncertain.
A November 12 report from ThorstenMeyerAI.com says AI has delivered measurable productivity gains in some individual workplaces, but its contribution to overall productivity growth remains modest so far. The report cites a study finding a 15% gain for customer-support agents and says executives put AI’s current contribution to productivity growth at roughly 0.2 percentage points a year—a distinction that matters as businesses and policymakers judge whether early results signal a broader economic shift.
The report frames productivity as output per hour worked, or the amount of value produced for each hour of labor. It argues that small differences in the annual growth rate compound over time: productivity growing at 2% a year takes about 35 years to double, while growth at 1% takes about 70 years. These are illustrative calculations, not forecasts of how quickly AI will raise productivity.
To show why technology alone does not guarantee immediate economywide gains, the report revisits economist Robert Solow’s 1957 analysis of US output per worker between 1909 and 1949. Solow attributed 87.5% of the increase to technical change and 12.5% to increased use of capital. The residual category can include improved methods, organization and skills, not only new machines.
The report also compares current AI claims with the computer era. US productivity growth slowed after 1973 even as computers spread, then accelerated to about 3.1% a year from 1995 to 2004, based on the report’s calculation from a Bureau of Labor Statistics index. It cites Federal Reserve economists’ estimate that information technology accounted for about two-thirds of the one-percentage-point acceleration in the late 1990s. The historical sequence shows a lag can occur, but does not establish that AI will produce the same pattern.
Infographic 01 · The sector clock
Productivity moves the sector clock
US farms produced nearly three times their 1948 output in 2023, with only about 9% more inputs in total. Over the same decades, farming's share of US jobs fell from about one in eight to 1.5%.
- Field12.1%
- Factory29.9%
- Services58.0%
Lebergott · own calculation
- Field1.5%
- Factory18.9%
- Services79.6%
ILO estimate
Farm output +210% (1948-2023, 1.52% a year); all inputs +0.12% a year (about +9% in total); total factor productivity +1.40% a year (USDA ERS). 1950: services = remainder, incl. unemployed.
Sources: USDA ERS, Agricultural Productivity in the U.S. (updated 25 Aug 2026); US employment shares: S. Lebergott, in US Census Bureau, Historical Statistics of the United States (1975; 1950, own calculation, see episode 2); ILO modelled estimates via World Bank (2025).
Powered by Thorsten Meyer AI
Why Economywide Gains Take Time
The central point for workers, companies and governments is that local efficiency gains are not the same as economywide productivity growth. A tool can help a team handle more work in less time, yet the effect on national output per hour depends on how widely it is adopted, whether workflows change, and whether the gains persist after implementation.
Productivity growth affects how much an economy can produce without relying only on longer hours or a larger workforce. It can support rising living standards over time, although the distribution of any gains is a separate question. The report’s historical examples also caution against treating impressive demonstrations or isolated studies as proof of a broad economic transformation.
For employers, the distinction affects how they assess AI investments. A productivity claim should be tested against a clear baseline and a defined time period, and should account for implementation costs and changes in work quality. The report does not provide enough detail about the cited support-agent study or the executive estimate to determine how directly either measure captures those factors.
Popular right now
Bose QuietComfort Wireless Noise-Cancelling Over-Ear Headphones, Twilight Blue
LG StanbyME 2 Max 32-Inch 4K Portable Smart Touchscreen TV
Ninja Luxe Café Mini Plus Espresso and Coffee Maker with AutoFroth, Bone & Champagne
As an affiliate, we earn on qualifying purchases.
From Computers to AI
The computer era provides the report’s main historical comparison. In a 1987 review, Solow observed that the perceived technological revolution had coincided with slower productivity growth. His line—“You can see the computer age everywhere but in the productivity statistics”—became known as the Solow paradox. Productivity later accelerated in the late 1990s, though that chronology alone cannot prove computers caused the change.
The report also uses US agriculture to illustrate how productivity changes the mix of work. It says farm output in 2023 was nearly three times its 1948 level while total inputs rose by about 9% across that period, citing US Department of Agriculture data. It places this alongside the decline in agriculture’s share of US employment, from roughly one in eight workers in 1950 to an estimated 1.5% in 2025. These figures show how output can rise even as fewer people work in a sector; they are not a direct measure of AI’s effects.
The historical productivity figures in the report refer to specific measures and periods. Its postwar and computer-era growth rates use a BLS index for the nonfarm business sector, while Solow’s earlier calculation concerns output per man-hour. Comparisons across those measures should not be treated as though they describe one uninterrupted series.
“You can see the computer age everywhere but in the productivity statistics.”
— Robert Solow, quoted in a 1987 New York Times Book Review review
Limits of the Early AI Estimates
The source does not identify the customer-support study by title, author, sample size, publication status or measurement period. It is therefore not possible from the supplied material to assess how the 15% figure was calculated, whether it represents a measured increase in output per hour, or how broadly it applies to other jobs and companies.
The report attributes an estimate of roughly 0.2 percentage points a year in productivity growth to executives, but it does not specify the survey, respondent group, time window or method behind that figure. The estimate should be read as an attributed assessment, not as a settled official statistic. The source also does not establish whether the reported gains persist, offset the cost of adopting AI, or improve the quality of work.
It remains unclear when AI might affect national productivity measures at scale. Adoption, worker training, investment, business-process changes and measurement practices could all shape the timing and size of any future effect. The computer-era comparison offers context, not a forecast.
Evidence to Watch in Coming Data
The next test will be whether gains reported in particular workplaces appear consistently across industries and over longer periods. Readers can watch for studies that disclose their methods and comparison groups, alongside official productivity data that specify the sector, measure and time frame. Those sources can help distinguish isolated improvements from a broad increase in output per hour.
Companies assessing AI should report what work was measured, what baseline was used, and whether time saved was converted into additional output or better service. At the economywide level, forthcoming productivity releases will show whether measured growth changes, but they will not by themselves identify AI as the cause. The source offers no specific date for a next report or data release tied to its estimates.
Key Questions
What does productivity mean in this report?
It means output per hour worked: how much value workers produce for each hour of labor. The report focuses on labor productivity and also discusses total factor productivity, a measure of output growth not explained by increased labor and capital inputs.
Has AI raised productivity by 15%?
The source says one study found a 15% gain among customer-support agents. It does not provide the study’s full details, and the result should not be treated as a general estimate for all workers or the economy.
What is AI’s current economywide productivity effect?
The report says executives estimate AI has contributed about 0.2 percentage points a year to productivity growth so far. It does not identify the survey or methodology, so the figure is an attributed estimate rather than a confirmed official measure.
Why did computers not immediately raise productivity statistics?
Solow noted in 1987 that computers were visible across workplaces while productivity growth had slowed. Growth later accelerated in the late 1990s, but the historical timing does not prove that technology alone caused the increase or predict what will happen with AI.
When will AI’s effect become clear in national data?
The source gives no timetable. It remains uncertain how quickly adoption and changes in work practices will show up in economywide measures, or how much of any future change can be attributed specifically to AI.
Source: ThorstenMeyerAI.com
