🔍 Read the full analysis: What Can We Learn From AI Metrics? 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
ThorstenMeyerAI.com’s Nov. 11 report uses the history and limitations of GDP to examine what economic metrics reveal—and omit—as AI tools spread. It cites a Stanford working paper estimating $172 billion a year in value to U.S. consumers from AI chatbots, while stressing that consumer value, company revenue and broader welfare are different measures.
A Nov. 11 report from ThorstenMeyerAI.com argues that the history of GDP offers a warning for anyone trying to measure artificial intelligence: economic statistics can capture market activity while missing unpaid work and much of the value people receive from free services. The report cites a Stanford working paper estimating that U.S. consumers receive $172 billion a year in value from AI chatbots, an estimate that is not the same as the tools’ revenue or a measure of their overall effect on welfare.
The report traces national income measurement to the Great Depression. After the U.S. Senate requested estimates, economist Simon Kuznets led work that produced a report delivered on Jan. 4, 1934. It estimated that income produced in the United States fell from $83.0 billion in 1929 to $39.4 billion in 1932, a 53% decline. The figures gave lawmakers a way to gauge the scale of the contraction at a time when comprehensive national measures were lacking.
Kuznets also warned that national income was not a direct measure of public welfare. The early estimates focused mainly on activity represented in markets, excluding much household labor. They did not show how income was distributed or account for how difficult it was to earn. The report quotes his conclusion that “the welfare of a nation can, therefore, scarcely be inferred from a measurement of national income.”
The article applies that limitation to digital services and AI. It points to a 2017 experiment in which the median participant would have required $17,530 to give up search engines for a year, despite paying no direct price for them. It also cites a Stanford working paper’s estimate of $172 billion in annual U.S. consumer value from AI chatbots. The source does not provide the paper’s methods or publication status in the supplied material, so the figure should be read as an attributed estimate, not an established accounting total.
Infographic 01 · The sector clock, by output
China: how GDP changed shape
Share of GDP (value added at current prices) by sector, 1978 to 2025. In 2023 farming still held 22.8% of all jobs but produced 7.1% of GDP.
- Farming27.7%
- Industry47.7%
- Services24.6%
Reform starts
- Farming14.7%
- Industry45.5%
- Services39.8%
- Farming9.1%
- Industry45.4%
- Services45.5%
Services overtake industry
- Farming6.7%
- Industry35.6%
- Services57.7%
Communiqué 2025
Measured by output, not by jobs: the sector clocks of earlier episodes count workers. Jobs vs. output for farming, 2023: 22.8% of employment (year-end) vs. 7.1% of GDP.
Sources: NBS China, China Statistical Yearbook 2024, tables 3-2 and 4-2; NBS, Statistical Communiqué 2025 (28 Feb 2026).
Powered by Thorsten Meyer AI
AI Value Beyond Company Revenue
The report’s central implication is that revenue and consumer benefit are not interchangeable. A service can generate substantial value for users without charging them an equivalent amount, while a large market or a high valuation does not by itself establish broad social benefit. Measures based mainly on market transactions can therefore understate some benefits and fail to describe who receives them.
This distinction matters as businesses and policymakers evaluate AI’s economic contribution. A chatbot’s estimated value to users does not establish how much it raises productivity, how benefits are distributed, or whether gains outweigh costs such as labor disruption, errors or infrastructure spending. Those questions require evidence beyond a single headline figure. The report makes a case for reading AI statistics alongside their definitions, methods and comparison periods—not treating any one measure as a complete account.
Products Worth Considering
Bose TV Speaker Compact Soundbar with Bluetooth and HDMI-ARC, Black
DJI Mini 4 Pro Fly More Combo Drone with DJI RC 2
Sony WH-1000XM5 Premium Noise Cancelling Wireless Headphones – Midnight Blue
As an affiliate, we earn on qualifying purchases.
From Depression-Era Accounts to GDP
National accounts developed in response to the need for a clearer picture of economic activity. The report says the United States expanded estimates of gross national product during the 1940s to help plan wartime production. The first international System of National Accounts followed in 1953, and in 1991 the United States adopted gross domestic product as its headline measure, in line with international standards.
GDP counts the market value of final goods and services produced within a country. In the U.S. accounting framework, it is calculated from consumption, investment, government spending and net exports. That definition is useful for tracking production, but it does not make GDP a universal measure of quality of life. The source cites Germany’s unpaid-work figures as an example: adults aged 18 and older averaged 24 hours and 58 minutes of unpaid work a week in 2022, compared with 20 hours and one minute of paid work. Such work matters to households even when it is not counted as market output.
The report also gives a historical illustration of how the composition of production changes: services rose from 24.6% of China’s value added in 1978 to 57.7% in 2025, while farming’s share fell. These are shares of output, not shares of workers. That difference reinforces the need to specify what a metric measures before drawing conclusions about jobs or living standards.
“The median user would have needed $17,530 to give up search engines for a year.”
— ThorstenMeyerAI.com, describing a 2017 experiment
What the Chatbot Estimate Measures
The supplied source identifies the $172 billion figure as an estimate from a Stanford working paper, but it does not name the paper, its authors, its methods or the date of the estimate. It is not clear from the material how researchers defined consumer value, which chatbot services or users were included, or how uncertainty was calculated. The source also does not give a comparable revenue figure, so the size of any gap cannot be established from the information provided.
More broadly, the report does not quantify AI’s effects on productivity, employment, costs or public welfare. The consumer-value estimate should not be treated as proof that AI has produced a particular net economic gain. The period covered, treatment of free or bundled services, and distribution of estimated benefits remain unspecified in the supplied text.
Better Measures for AI’s Effects
The immediate next step for readers assessing the claim is to consult the cited Stanford working paper for its authors, methods, assumptions and time frame. Further evidence would be needed to compare estimated user value with company revenue and costs on consistent terms, and to determine how benefits vary across users and sectors.
As AI adoption continues, economic assessments will need to distinguish production, prices, consumer surplus, productivity and distribution rather than combining them under a single label. The report supplies a historical caution, not a final measure of AI’s contribution: statistics can be useful and still leave important parts of economic life out.
Key Questions
What is the $172 billion figure?
It is a Stanford working paper estimate, cited by ThorstenMeyerAI.com, of annual value to U.S. consumers from AI chatbots. The supplied material does not include the paper’s methods or time frame details.
Does the estimate mean chatbot companies earn $172 billion?
No. The source describes the number as consumer value, not company revenue. The two measures capture different things, and the supplied text gives no directly comparable revenue total.
Why does the report discuss GDP?
GDP is a way to track market production, but it does not capture every aspect of welfare. The report uses its history to show why AI’s economic effects may also require measures of user benefit, unpaid activity and distribution.
What did Simon Kuznets warn about?
Kuznets cautioned that national income figures alone could not establish a nation’s welfare. The early accounts left out much unpaid household work and did not show how income was shared.
Source: ThorstenMeyerAI.com
