🔍 Read the full analysis: AI's Disruption Of Knowledge Work on ThorstenMeyerAI.com
Get business pricing on tech for your team
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
TL;DR
A November 10, 2026, report from ThorstenMeyerAI.com revisits how knowledge and information work became a major part of the U.S. economy, then connects that history to generative AI’s focus on office tasks. It cites estimates of high exposure for clerical and higher-income work, but stresses that exposure is not the same as job loss and that broad U.S. employment data has not yet shown a clear AI effect.
ThorstenMeyerAI.com published a report November 10 tracing the rise of knowledge and information work and examining how generative AI may affect it. The report says clerical and higher-income jobs appear especially exposed to the technology, while cautioning that exposure does not establish that jobs will be eliminated and that U.S. employment data has not yet shown a clear overall AI footprint.
The report grounds the discussion in two historical estimates. Economist Fritz Machlup calculated that knowledge production and distribution accounted for $136.4 billion, or “almost 29 per cent of adjusted GNP,” in the United States in 1958. His estimate measured spending on knowledge-related activity, not the value of knowledge itself. The report says education made up the largest part of the total, followed by media, information services, research and development, and information machines.
It also draws on Marc Porat’s 1977 study for the U.S. Department of Commerce, which classified workers by the tasks they performed rather than by their employer’s industry. Porat’s figures put information workers at 5.8% of the U.S. workforce in 1860, 30.8% in 1950 and 46.4% in 1970. The report says information work had become the largest of the four groups—agriculture, industry, services and information—by the mid-1950s.
For a current comparison, the report 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. Those occupational categories indicate the scale of knowledge-related employment, but they are not themselves a measure of AI-driven job losses.
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
Office Work Faces AI Exposure
The development matters because generative AI is aimed at tasks that make up a large share of knowledge work: writing, reading, planning, calculating and handling information. The historical figures in the report show that this is not a small occupational niche. Information-related work grew into a major part of the U.S. workforce over more than a century, and the cited present-day figures put managers, professionals and technicians at substantial shares of employment in three economies.
But the distinction between task exposure and employment outcomes is central. A tool may assist with, speed up or change a task without removing the job that contains it. The report says overall U.S. job data has not yet revealed a clear AI effect, while pointing to evidence that younger workers in the most exposed jobs are falling behind their peers. That observation is a warning sign, not proof that AI caused the difference; the report does not provide enough detail here to establish a causal link.
For workers, employers and policymakers, the immediate question is not simply how many jobs can use AI. It is how organizations deploy it, which tasks change, and whether those changes affect hiring, pay, productivity or job security. The available indicators described in the report do not settle those questions.
Products Worth Considering
Samsung Galaxy Tab A11+ 8GB RAM, 256GB Storage, Gray
DJI Osmo Mobile 8 3-Axis Phone Gimbal Stabilizer
LG UltraGear 27-inch QHD OLED Gaming Monitor, 240Hz
As an affiliate, we earn on qualifying purchases.
From Knowledge to Information Work
The report places today’s AI debate within a longer shift in how economies are organized. In 1959, management writer Peter Drucker described the rise of “knowledge work”, arguing that more people would earn a living through what they had learned rather than primarily through physical labor. The report says this shift would require different management practices from those developed for factory production.
Machlup tried to quantify the emerging activity in his 1962 book, The Production and Distribution of Knowledge in the United States. He counted the costs of industries that produced or distributed knowledge, including education, research, media, information machines and information services. This approach offered an economic estimate, but it had a limit: spending is not the same as knowledge’s value.
Porat later focused on occupations and tasks, including information workers employed outside sectors usually associated with information. That method produced a different measure from Machlup’s industry spending. The report’s historical account therefore describes related but distinct ways of counting the economy’s information activity—not a single directly comparable series.
““almost 29 per cent of adjusted GNP””
— Fritz Machlup, as quoted in the report
Exposure Is Not Job Loss
The scale and timing of AI’s labor-market effects remain unclear. The report does not provide a quantified estimate of jobs already eliminated by generative AI, nor does it identify a clear overall AI footprint in U.S. employment data. Its statement about younger workers in highly exposed occupations falling behind peers is not accompanied in the supplied material by a dataset, time period or causal analysis sufficient to show that AI produced the difference.
It is also unclear how the exposure estimates define a job as exposed, which AI systems or tasks they cover, and how quickly employers will adopt those systems. The figures on managers, professionals and technicians describe occupational shares, not the portion of work automated. Historical estimates from Machlup and Porat use different methods and should not be treated as direct forecasts of AI’s effects.
Track Hiring and Task Changes
The report presents no specific employment forecast or scheduled policy decision. The next meaningful evidence will come from continued labor-market data and workplace studies that track hiring, wages, hours, productivity and task composition over time, especially in occupations with high measured AI exposure. Those indicators can help distinguish changes associated with AI from broader economic shifts.
Employers’ deployment decisions will also shape the outcome: AI may be used to support workers, reorganize duties or reduce demand for some tasks. Until evidence links those choices to measurable employment changes, the report’s central conclusion remains bounded: knowledge work is a large and plausible target for generative AI, but its net effect on jobs is not yet established.
Key Questions
What is the news development?
ThorstenMeyerAI.com published a report on November 10, 2026, connecting the historical growth of knowledge and information work with current questions about generative AI’s effects on office jobs.
How much of the U.S. economy did Machlup attribute to knowledge activity?
Machlup estimated that knowledge production and distribution represented $136.4 billion, or almost 29% of adjusted U.S. GNP, in 1958. The figure measured costs and spending, not the value of knowledge itself.
Does AI exposure mean a job will be lost?
No. Exposure means tasks may be affected by AI; it does not by itself show that an employer will eliminate a role. The report says broad U.S. employment data has not yet shown a clear overall AI footprint.
What evidence does the report cite about current employment?
It cites 2025 ILO occupational shares: managers, professionals and technicians held 62% of jobs in Singapore, 52.5% in the United States and 47.9% in Germany. These figures show the size of those occupational groups, not AI-caused job losses.
Source: ThorstenMeyerAI.com
