🔍 Read the full analysis: AI Metrics: What They Show And What They Leave Unsaid on ThorstenMeyerAI.com
Get the little things that make your day 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 article examines what GDP measures and what it omits, using its history and examples including unpaid work and free digital services. It cites a Stanford working paper estimating $172 billion a year in consumer value from U.S. AI chatbots, but the estimate is not the same as revenue or a complete measure of AI’s social effects.
ThorstenMeyerAI.com published an article on Nov. 11 arguing that familiar economic measures such as GDP can miss value that has no market price, including unpaid household labor and consumer benefits from free digital tools. The article cites a Stanford working paper estimating $172 billion a year in value to U.S. consumers from AI chatbots, while emphasizing that this estimate is not the revenue those services earn and does not settle what AI contributes to broader welfare.
The article begins with the creation of a national income measure during the Great Depression. In 1932, the U.S. Senate asked the Commerce Department to estimate national income. Economist Simon Kuznets supervised the work, and the resulting report, National Income, 1929–32, was submitted on Jan. 4, 1934. It estimated that income produced fell from $83.0 billion in 1929 to $39.4 billion in 1932, a decline of 53%. Income paid to individuals fell by 40% over the same years.
Kuznets cautioned that the figures mainly covered activity whose results appeared in markets. His report excluded much household work and did not show how income was distributed or how much effort earning it required. The article quotes his warning that “the welfare of a nation can, therefore, scarcely be inferred” from national income alone. The measure provided a way to track economic output, not a complete account of how people were doing.
The article then applies that distinction to contemporary measures. It says unpaid household work in Germany was valued at €987 billion in 2013 using wage estimates, equivalent to 39% of the value added counted in GDP. It also cites 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 to use them. These examples illustrate the gap between market spending and estimated consumer benefit.
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
Why AI Value Can Outrun Revenue
The comparison matters because AI services can create value for users without capturing all of it as sales. A chatbot may help someone draft text, find information or complete a task, while the user pays nothing or less than the estimated value of that help. Revenue and consumer benefit answer different questions: revenue records what providers earn, while consumer-surplus estimates attempt to measure how much users value access beyond what they pay.
The article’s cited estimate of $172 billion annually for U.S. consumers could inform debate about AI’s economic contribution, but it should not be read as a forecast, a measure of company income or proof of net social benefit. A measure of perceived benefit does not automatically account for costs such as subscriptions, computing resources, errors, privacy risks or displaced work. The number is an estimate of one category of consumer value, not a full balance sheet for AI.
The broader point reaches beyond AI. GDP remains useful for tracking market production and comparing economic activity, but treating it as a direct score of well-being can leave important activity and consequences out of view. What a metric counts shapes what decision-makers can see; readers need to know its scope before drawing conclusions from it.
Products Worth Considering
Withings Body Scan Smart Scale with Segmental Body Composition and ECG, Black
Garmin Forerunner 265S 42 mm GPS Running Watch, Light Pink and Powder Gray
Philips Sonicare ProtectiveClean 5300 Rechargeable Electric Toothbrush with 3 Brush Heads, Black
As an affiliate, we earn on qualifying purchases.
From Depression-Era Accounts to GDP
The 1934 national-income report was an early attempt to quantify the scale of an economic collapse when policymakers had limited data. The article traces later changes: wartime planning brought estimates of gross national product, an international system of national accounts appeared in 1953, and the United States adopted GDP as its headline measure instead of GNP in 1991. GDP counts the market value of final goods and services produced within a country, commonly expressed through consumption, investment, government spending and net exports.
That design makes GDP valuable for tracking production, but it does not count every activity that supports people’s lives. The source gives unpaid care and housework as examples, and describes free digital services as another case where use may be valuable even without a market payment. The article also notes that output and employment can tell different stories: in China, farming accounted for 6.7% of value added in 2025, while the source says agriculture employed 22.8% of workers at the end of 2023. Those figures refer to different years and measures, and show why output statistics do not describe how work is distributed.
“The estimates “cover primarily only efforts whose results appear on the market place of our economy”.”
— Simon Kuznets, in the 1934 report, as quoted by ThorstenMeyerAI.com
What the Chatbot Estimate Leaves Out
The source identifies the Stanford figure as an estimate but does not provide the working paper’s title, authors, publication date, methods or uncertainty range. It is therefore not possible from the supplied material to assess how researchers valued chatbot use, which products or users were included, or how the estimate handles differences in service quality and frequency of use. The article’s comparison with company revenue is also not accompanied here by a specific revenue total or a shared measurement period.
It is also unclear how much of the estimated value reflects benefits that users would retain if a particular chatbot were unavailable, or how the estimate accounts for costs and harms. The figure should be understood as a source-attributed estimate of consumer value, rather than a confirmed accounting of AI’s net effect. The source material supplied for this article ends during its discussion of unpaid work, so further details from that section cannot be verified here.
Scrutinize the Measures and Methods
The next step for readers is to consult the Stanford working paper and the underlying national-statistics sources to check definitions, methods and dates. Any comparison between chatbot value and provider income should use figures that cover the same geography and period and clearly distinguish consumer surplus from revenue. Policymakers and businesses using AI metrics should also state what is counted and what is left outside the calculation.
ThorstenMeyerAI.com presents the article as part of its “Work After Work” series, with later installments planned to examine productivity and other limits of economic measures. For now, the reported chatbot figure is an estimate cited by the article; its precise methodology and implications remain open to scrutiny.
Key Questions
What is the $172 billion AI figure?
It is a Stanford working-paper estimate, cited by ThorstenMeyerAI.com, of annual value to U.S. consumers from AI chatbots. It is not presented as chatbot-company revenue.
Why does GDP miss unpaid household work?
GDP primarily measures market production. Work such as cooking, cleaning and caregiving often has no market transaction, so it may not appear in GDP even though it supports households and communities.
Does a high consumer-value estimate prove AI improves welfare?
No. An estimate of how much consumers value a service does not by itself account for all costs, risks, distributional effects or impacts on workers. It is one measure, not a complete welfare assessment.
What did Kuznets say about national income?
In the 1934 report, Simon Kuznets warned that national income should not be treated as a direct measure of welfare. He noted that the estimates focused mainly on activity reflected in markets and did not capture several dimensions of people’s lives.
Source: ThorstenMeyerAI.com
