🔍 Read the full analysis: AI Driven Changes In Knowledge Work on ThorstenMeyerAI.com
Prime made for students and young adults
- Fast, free delivery for dorm and study essentials
- Prime Video and Amazon Music included
- Member-only deals
TL;DR
A November 10, 2026, article from ThorstenMeyerAI.com traces the growth of knowledge and information work, from Fritz Machlup’s estimate that it accounted for nearly 29% of adjusted U.S. GNP in 1958 to its later expansion into a major employment sector. It says generative AI is aimed at tasks common in office jobs, but exposure does not establish that jobs will disappear; broad U.S. employment data has not yet shown a clear AI effect.
Knowledge work now makes up a large share of employment in wealthy economies, and the office tasks central to it are among those most exposed to generative AI, according to a report published November 10 by ThorstenMeyerAI.com. The report traces that workforce shift through economic studies dating to the 1950s and cautions that exposure to AI is not the same as job loss; broad U.S. employment data has yet to show a clear AI footprint.
The report begins with economist Fritz Machlup’s 1962 study, which estimated that spending on producing and distributing knowledge in the United States reached $136.4 billion in 1958, or “almost 29 per cent of adjusted GNP.” The figure measured costs, not the economic value of knowledge itself. Education accounted for the largest share of the total, followed by media, information services, research and development, and information machines.
It also describes Marc Porat’s later classification of workers by the tasks they performed, rather than the industry where they worked. In his 1977 report, Porat estimated that information workers made up 46.4% of the U.S. workforce in 1970, compared with 5.8% in 1860. The article says information work had become the largest of four workforce groups by the mid-1950s, ahead of agriculture, industry and services.
For a contemporary comparison, the source cites 2025 International Labour Organization figures showing managers, professionals and technicians held 62% of jobs in Singapore, 52.5% in the United States and 47.9% in Germany. It says higher-income and clerical occupations are at the front of the line for generative AI exposure. That describes potential contact with the technology, not confirmed employment losses or a measured number of jobs replaced.
Infographic 01 · The number to remember
Fritz Machlup's first count of the knowledge industries, by branch.
US spending on knowledge by branch, 1958 · share of $136.4 bn
Machlup measured what knowledge cost (spending, the way national accounts count), not what it was worth. “Almost 29 per cent of adjusted GNP” is his figure for 1958. The branch values are as tabulated by Godin (2008); the labour shares are Machlup’s estimate as summarised by his publisher.
Sources: F. Machlup, The Production and Distribution of Knowledge in the United States (Princeton University Press, 1962), as tabulated by B. Godin, “The Knowledge Economy: Fritz Machlup's Construction of a Synthetic Concept” (2008); Princeton University Press, book description.
Powered by Thorsten Meyer AI
AI Reaches the Office Workforce
The shift matters because generative AI is being applied to activities that define much knowledge work: writing, reading, planning, checking and handling information. Those tasks are not confined to technology companies. They appear across professional, managerial and clerical roles, meaning potential changes could affect a broad share of workers and employers.
But potential exposure leaves open what happens to those jobs. AI could change how tasks are done without eliminating the roles that contain them; the source does not provide evidence establishing how much work has been automated or how productivity and wages have changed as a result. It reports that overall U.S. job data shows no clear AI footprint so far, while younger workers in the most exposed jobs are falling behind their peers. That reported pattern warrants attention, but by itself does not prove AI caused the difference.
The historical account also puts the current debate in perspective: information work became a major part of the economy long before generative AI. The present question is not whether knowledge work exists, but how AI may alter the tasks, skills and opportunities within it.
Popular right now
Ninja CREAMi Scoop & Swirl 13-in-1 Ice Cream and Soft Serve Maker, NC701
Ninja Crispi Microwave and Air Fryer with FusionCrisp, Silver, MW101
Samsung 43-Inch M70H Series 4K Mini LED Smart TV (2026)
As an affiliate, we earn on qualifying purchases.
From Knowledge Work to AI
Management writer Peter Drucker named “knowledge work” in 1959, describing workers who earn a living through what they have learned rather than primarily through manual labor. He argued that factory management principles would not automatically suit people whose work depends on expertise and judgment. The source presents this as an early account of a growing occupational shift, not as a prediction about today’s AI systems.
Machlup then tried to quantify the knowledge economy by measuring expenditure across education, research and development, communications media, information machines and information services. His estimate of nearly 29% of adjusted GNP in 1958 was a measure of the cost of producing and distributing knowledge. Porat’s later approach tracked information-related occupations and industries, offering a different measure. The two figures describe different things and should not be treated as interchangeable measures of the same economic share.
The article’s broader frame is a move from agriculture and manufacturing toward services and information work. Its account links that long-running change to the current spread of AI into office tasks, while leaving the size and consequences of that newer change unsettled.
““Exposure is not job loss, yet.””
— ThorstenMeyerAI.com report
What AI Exposure Does Not Show
The source does not specify the methodology, scope or time period behind its statement that higher-income and clerical jobs are in front of the line for generative AI exposure. It also does not give a numerical estimate of jobs likely to be displaced, or establish how many workers’ tasks have already changed because of AI. Exposure estimates are not forecasts of job losses.
The article says U.S. job data shows no clear AI footprint so far, but provides no dataset, measurement window or comparison baseline for that statement. It also reports that young workers in the most exposed jobs are falling behind peers without identifying a specific measure or demonstrating that AI caused the gap. The timing, scale and distribution of any future employment effects remain unclear.
Tracking Work as AI Spreads
The next evidence to watch is whether sustained labor-market data shows changes in hiring, employment, wages or job duties in occupations that use generative AI. Those measures would help distinguish between tools that alter tasks and tools associated with reduced demand for workers. The source does not announce a specific new dataset, study or policy milestone.
ThorstenMeyerAI.com frames this report as the final episode of its first arc on how the economy is built and says the next episode will examine what economic statistics measure and what they leave out. For the AI question, the central unresolved test is whether exposure translates into observable changes—and, if it does, which workers and tasks are affected.
Key Questions
What is knowledge work?
Knowledge work is work that relies primarily on learned information and expertise. The source associates it with tasks such as planning, reading, writing, calculating and managing information.
What did Machlup’s 29% estimate measure?
Machlup estimated that spending on producing and distributing knowledge in the United States was $136.4 billion in 1958, or almost 29% of adjusted GNP. The estimate measured expenditure, not the value of knowledge itself.
Does AI exposure mean a job will disappear?
No. Exposure indicates that work may involve tasks AI could affect; it does not establish that an entire job will be automated or eliminated. The source reports no clear overall AI footprint in U.S. job data so far.
How large is information work in the figures cited?
Marc Porat estimated that information workers made up 46.4% of the U.S. workforce in 1970. The source also cites 2025 figures for managers, professionals and technicians: 52.5% of jobs in the United States, 62% in Singapore and 47.9% in Germany.
Source: ThorstenMeyerAI.com
