🔍 Read the full analysis: Understanding AI’s Measures Without Overreading Them 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
A November 11 report from ThorstenMeyerAI.com uses the history of national income statistics to argue that economic measures should not be mistaken for measures of welfare. It cites estimates of unpaid work and AI chatbot benefits to show how value can be missed or counted differently, while noting that such estimates do not by themselves establish broad economic impact.
ThorstenMeyerAI.com published an analysis on November 11 arguing that figures used to measure economic output should not be read as complete measures of well-being or the value people get from artificial intelligence. The report links economist Simon Kuznets’ cautions about national income statistics to recent estimates of unpaid household work and consumer benefits from AI chatbots.
The analysis begins with a report presented to the U.S. Senate on January 4, 1934. Planned and supervised by economist Simon Kuznets, it estimated that U.S. national income produced fell from $83.0 billion in 1929 to $39.4 billion in 1932, a 53% decline. The figures gave lawmakers a way to quantify the Depression-era contraction, but Kuznets warned that they covered mainly work whose results appeared in markets.
Kuznets’ report excluded much household labor and did not show how income was distributed or how difficult the work was. It cautioned that national welfare could not be inferred directly from national income. The article uses that warning to frame its discussion of GDP, which measures the market value of final goods and services produced within a country and is commonly calculated from consumption, investment, government spending and net exports.
To illustrate what market-based measures can miss, the article cites German figures valuing unpaid household work at €987 billion in 2013, or 39% of the value added counted in GDP. It also refers to a 2017 study 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. These measures describe different kinds of value and should not be treated as directly interchangeable with GDP.
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
What GDP Leaves Outside the Market
The distinction matters because a measure can be useful without measuring everything people value. GDP helps governments, researchers and businesses track production and compare changes over time. It does not, by itself, describe distribution, unpaid contributions, environmental costs, leisure or whether people feel better off. Kuznets’ warning, as recounted in the article, is a reminder against turning an output statistic into a verdict on social welfare.
The same caution applies to claims about AI. A model or service can create benefits that are not reflected in its sales, but a calculated consumer benefit is not the same as company revenue, measured productivity growth or a demonstrated improvement in living standards. Readers and policymakers need to know what is being measured, how the estimate was made and what comparison it uses before drawing conclusions about AI’s economic effects.
This also affects public debate over whether AI is already delivering major economic gains. If analysts focus only on spending and business income, they may miss consumer benefits; if they rely only on estimated willingness to pay, they may overstate what has been established about economy-wide change. Keeping the measures distinct makes comparisons more informative.
Products Worth Considering
Amazon Kindle Paperwhite (2024, 16 GB) – 7-inch Glare-Free Display, Black
Samsung Galaxy Tab A11+ (6GB RAM, 128GB Storage) – Gray
DJI Mini 4K Drone with 4K UHD Camera
As an affiliate, we earn on qualifying purchases.
From Depression Statistics to GDP
The 1934 national-income report emerged after the Senate asked the Commerce Department for estimates during the Great Depression. Before a comprehensive series existed, policymakers relied on partial indicators, including stock prices and industrial production. Kuznets’ estimates supplied a clearer account of the fall in income, while their stated limits helped establish an important distinction between measuring production and judging welfare.
The U.S. later developed gross national product estimates to support wartime planning. The United Nations’ first international System of National Accounts appeared in 1953, and the United States adopted GDP as its headline measure in 1991. GDP’s role is to track production within a country, not to serve as a complete score of quality of life. The article notes that U.S. GDP was about $30.86 trillion in 2025, but that scale alone does not answer how income or benefits are distributed.
For its AI comparison, the article cites a Stanford working paper estimating $172 billion a year in consumer value from U.S. AI chatbots. That is an estimate of consumer benefit, not a reported amount of sales or a national-accounts statistic. The article’s larger point is methodological: economic measures can vary according to whether they record market transactions, household production or estimated value to users.
“The estimates cover primarily only efforts whose results appear on the market place of our economy.”
— Simon Kuznets, in the report
Limits of the AI Value Estimate
The source describes the $172 billion figure as an estimate in a Stanford working paper; it does not provide the paper’s methods, assumptions, publication status or uncertainty range. The figure should not be presented as confirmed revenue, measured productivity or a settled account of AI’s contribution to the U.S. economy. It is also unclear from the supplied material how the estimate compares with earlier years or alternative methods.
The article gives a willingness-to-accept figure for search engines and a separate estimate for chatbot benefits, but those examples use different services and measurement approaches. The supplied text does not establish that either value is realized by every user, nor does it say how benefits vary among people. The German unpaid-work statistic also concerns a different year and accounting concept than GDP, so it illustrates a blind spot rather than providing a like-for-like adjustment to the national total.
Reading Future AI Statistics
The analysis does not report a new government rule, statistical release or scheduled research milestone. Its practical implication is that future AI claims should be read alongside the methods and definitions behind them. For the Stanford estimate, readers would need the underlying paper and its assumptions to judge how the $172 billion figure was calculated and how robust it is.
For policymakers and companies, the next step is to keep separate measures of sales, productivity, consumer surplus and unpaid or non-market activity, rather than treating one as a proxy for all the others. Further evidence on how AI tools affect actual work, prices and user outcomes will be needed to establish whether estimated consumer value corresponds to broader economic gains.
Key Questions
What is the main point of the analysis?
It argues that GDP measures market production, not overall welfare, and that estimates of AI’s value need to be interpreted according to what they measure.
What did Simon Kuznets measure in 1934?
The Senate report estimated U.S. national income produced, including a fall from $83.0 billion in 1929 to $39.4 billion in 1932. Kuznets also described important limits, including the exclusion of much unpaid household work.
Does the $172 billion AI figure mean chatbots generated that much revenue?
No. The article describes it as a working-paper estimate of annual consumer value from U.S. AI chatbots, not revenue earned by companies.
Why is unpaid household work discussed?
Much unpaid care and household labor is outside market transactions and therefore is not counted in GDP in the same way as paid services. The article cites a German valuation of €987 billion in 2013 to illustrate the scale of that excluded activity.
Source: ThorstenMeyerAI.com
