What AI Measures, From Performance To Progress
AIThis post was created with the assistance of artificial intelligence (AI).

🔍 Read the full analysis: What AI Measures, From Performance To Progress on ThorstenMeyerAI.com

Age 18–24?Offer from Amazon

Prime made for students and young adults

  • Fast, free delivery for dorm and study essentials
  • Prime Video and Amazon Music included
  • Member-only deals
Try Prime for Young Adults Free trial for eligible 18–24 year olds
As an affiliate, we earn on qualifying purchases.

TL;DR

ThorstenMeyerAI.com’s Nov. 11 report examines what GDP measures and what it leaves out, from unpaid household labor to free digital services and AI chatbots. It traces GDP’s origins in a 1934 Depression-era report and emphasizes economist Simon Kuznets’ warning that national income alone cannot measure a nation’s welfare.

ThorstenMeyerAI.com published a report on Nov. 11 examining how economic measures such as GDP capture market production while leaving out or understating unpaid work and the value people receive from free digital services, including AI chatbots. The report places today’s measurement debate in the history of a 1934 U.S. national-income report and its author Simon Kuznets’ warning that national income alone cannot establish a nation’s welfare.

The account begins with the Great Depression. In 1932, the U.S. Senate passed Resolution 220, asking the Commerce Department to estimate national income. Economist Simon Kuznets, working with the National Bureau of Economic Research, led the work. The resulting report, National Income, 1929–32, was submitted to the Senate 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 40%, from $81.1 billion to $48.9 billion.

The report says the early estimates mainly covered activity whose results appeared in markets. They left out much household work and did not show how income was distributed or how much effort earning it required. The article also describes later development of national accounts: the United States used gross national product estimates for wartime planning in the 1940s; the first international System of National Accounts appeared in 1953; and the U.S. adopted GDP as its headline measure in 1991.

To show the limits of market-based accounting, the report cites a German estimate valuing unpaid household work at €987 billion in 2013, or 39% of the value added counted in GDP. It also 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 the service. A Stanford working paper estimates that U.S. consumers receive $172 billion a year in value from AI chatbots; that is an estimate of consumer value, not chatbot revenue or a GDP figure.

At a glance
reportWhen: Published November 11, 2026
The developmentThorstenMeyerAI.com published a report connecting GDP’s origins and limitations with the difficulty of measuring unpaid work and consumer value from free AI tools.
Thorsten Meyer AI
Work After Work · Ep. 9

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.

Farming primary sectorIndustry incl. construction and miningServices tertiary sector
1978
  • Farming27.7%
  • Industry47.7%
  • Services24.6%

Reform starts

2000
  • Farming14.7%
  • Industry45.5%
  • Services39.8%
2012
  • Farming9.1%
  • Industry45.4%
  • Services45.5%

Services overtake industry

2025
  • 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

GDP’s Blind Spots in a Digital Economy

GDP remains a central measure of economic output, but it is not a complete account of living standards or the benefits people receive. Market prices make some activity straightforward to count, while unpaid care and household tasks generally do not appear as production in the same way. The distinction matters when policymakers and the public use output figures to describe how an economy is changing or how well people are doing.

The report’s examples also highlight a measurement challenge in digital markets. A service can provide substantial value to users while charging little or nothing directly. The Stanford estimate cited for AI chatbots concerns consumer benefit, while GDP records market production under its accounting rules. Those measures answer different questions. Treating an estimate of user value as revenue, or as a direct addition to GDP, would blur that distinction.

For readers, the practical point is not that GDP is useless, but that it is designed to measure a particular thing: production. The measure can help track economic change, as the Depression-era estimates did, while still failing to describe distribution, unpaid labor, or the full value users attach to free services.

As an affiliate, we earn on qualifying purchases.

From Depression Estimates to GDP

Before the 1934 report, U.S. leaders faced the Depression with incomplete indicators such as stock prices, freight-car loadings and limited measures of industrial production, according to the history summarized in the source. The Senate request produced a broader estimate of national income, giving officials a consolidated view of the collapse. Kuznets’ work was an early foundation for national accounting, not a claim that one figure could describe every dimension of economic welfare.

The accounting framework changed over time. The source says wartime planning drove further U.S. estimates of total output, followed by international standards and the 1991 switch from gross national product to GDP as the U.S. headline measure. GDP counts the market value of final goods and services produced within a country. In the report’s cited 2025 figures, U.S. GDP was about $30.86 trillion, Germany’s was €4.53 trillion and China’s was 140.2 trillion yuan. These are country-specific totals in different currencies, not directly comparable amounts without further adjustments.

The article also uses China’s output mix to illustrate structural change: services accounted for 57.7% of value added at current prices in 2025, compared with 24.6% in 1978. That output share is distinct from employment. The source says farming employed 22.8% of China’s workers at the end of 2023 while representing 7.1% of GDP, illustrating why output shares do not tell readers how many people work in each sector.

What the Estimates Do Not Settle

The source does not provide the full methods or underlying data for the German unpaid-work valuation, the 2017 search-engine experiment, or the Stanford working paper’s $172 billion AI estimate. The figures should be read as estimates with distinct methods and definitions, not as directly interchangeable measures. The report does not specify the working paper’s assumptions, uncertainty range, or how its consumer-value estimate relates to measured GDP.

It is also unclear from the supplied material how much of the estimated value of AI chatbots reflects services that would otherwise be unavailable, or how the estimate accounts for differences among users and products. The article’s broader point—that free services can deliver value not reflected in their price—is separate from whether any specific estimate is definitive. The source material ends during its discussion of German time-use data, so it does not establish the full set of figures or conclusions that may have followed.

Further Measures of Work and Value

The report frames unpaid work and digital services as measurement issues rather than proposing a replacement for GDP. The next step for readers assessing such claims is to examine the methods behind each estimate: what activity is counted, what comparison is used, and whether the result measures market output, time spent, or value to consumers.

The source also points toward a follow-up discussion of productivity, using the gap between China’s employment and output shares across sectors as a starting point. No publication date or additional findings for that discussion are provided in the supplied material. For now, GDP remains the central output measure described in the report, while the precise scale of excluded household activity and consumer benefits from free AI remains dependent on separate estimates.

Key Questions

What does GDP measure?

GDP measures the market value of final goods and services produced within a country during a given period. The report describes it as a measure of output, not a complete measure of welfare.

Why was the 1934 report significant?

It gave the U.S. Senate an estimate of the scale of the Depression-era decline: income produced fell 53% between 1929 and 1932, according to the report’s figures.

Does GDP count unpaid household work?

The early national-income estimates discussed in the article mainly covered activity with market results, leaving much household work outside their scope. The German figure cited for unpaid work is a separate valuation, not a standard GDP total.

Does the $172 billion AI estimate mean chatbots generated that much revenue?

No. The source describes the Stanford working paper’s figure as an estimate of annual value to U.S. consumers. It does not identify the amount as company revenue or GDP.

Source: ThorstenMeyerAI.com

You May Also Like

A Practical Guide To AI And Unemployment Categories

A ThorstenMeyerAI.com guide explains how official unemployment is counted and why AI-related job losses usually fall under structural unemployment.

How to Choose Science Reference Books For Students

Choose, check, and organize science reference books that fit a student’s age, curriculum, reading level, and research needs.

Terence Tao: Math 2.0 [Pdf]

A search-interest signal points to Terence Tao’s “Math 2.0” PDF, but the document’s contents and the reason for attention are unverified.

Best Science Reference Books For Students Compared

Compare general science encyclopedias with subject-specific reference books to choose the right format, depth, and value for a student’s needs.