For those who missed it, the U.,S. Council of Economic Advisors distributed, in January 2026, their Report: Artificial Intelligence and the Great Divergence; Press Release HERE). The analysis is grounded on a powerful analogy:
For centuries, most of the world’s economies grew at a similarly slow rate. However, a “Great Divergence” occurred with the Industrial Revolution, causing industrializing nations to accelerate their growth relative to the rest of the world. Artificial intelligence (AI) is a potentially transformative technology that is often compared to the Industrial Revolution. However, we are witnessing clear leaders in AI investment, performance, and adoption metrics across different nations. The Trump administration is laying the groundwork for American AI dominance by accelerating innovation, infrastructure development, and deregulation while establishing global dominance through technology exports. If the AI revolution is as transformative as the Industrial Revolution, should we expect this to lead to a second Great Divergence? (Artificial Intelligence and the Great Divergence)
The object, of course, is not merely dominance, but the protection of a space in which the state can further its ability to ensure the protection of the objectives of its political-economic model--managed protection of individual autonomous self fulfillment in and through markets based interactions in all spheres of social relations. To some significant extent its serves as the great power bookend to similar efforts to consider and situate the tech revolution within the cognitive cages of Chinese Marxist-Leninism and its political-social imperatives, which speaks in terms of high quality production fueling a socialist modernization that would push the nation farther along the Socialist Path.
At the same time, its consequences for the rest of the world can be understood as deeply divergent. That was the object of the UNDP Report (December 2025), The Next Great Divergence: Why AI May Widen Inequality Between Countries. This produces a counter analogy, one that focuses not on the consequences of transformative divergence among the great powers, but its effects on those beneath:
Artificial Intelligence is advancing rapidly, yet many countries remain without the infrastructure, skills, and governance systems needed to capture its benefits. At the same time, they are already feeling its economic and social disruptions. This uneven mix of slow adoption and high vulnerability may trigger a Next Great Divergence, where inequalities between countries widen in the age of AI. (The Next Great Divergence:)
One word, two meanings; one concept, two lenses; one evolutionary trajectory, two normative approaches; one factual reality, two analytical perspectives; one challenge, two policy responses. That divide, then, produces variation in values, signification, and approaches that will fundamentally shape policy and the legal structures within which states will seek to shepherd this challenge/opportunity in ways that advance their interests--or perhaps better put, the positive effects of these changes to their populations as a function of the goals and responsibilities of these states as measured against their ideologies.
 |
| Pix credit Lib Congress |
The AI revolution, with its parallels to the Industrial Revolution, presents a profound economic inflection point with the potential to significantly increase the GDP of countries that embrace it. We are witnessing clear leaders in AI investment, performance, and adoption metrics across different nations. The United States, as demonstrated by the comprehensive AI Action Plan and related executive orders from the Trump administration, is pursuing a strategy focused on accelerated innovation, infrastructure development, and establishing global dominance through technology exports and deregulation in order to lay the groundwork for American AI dominance. (Artificial Intelligence and the Great Divergence; p. 26).
The second is re-distributive, capacity sensitive, and risk averse--a cautious approach to technological revolution that value of which is measured against its costs, and the costs of which are measured against a set if value generating premises that tend to balance technological advancement against a specific set of human costs grounded in a specific set of adverse impact valuation measures. That is the essence of the UN approach in its The Next Great Divergence. But it also fuels middle power and Global South approaches There are variations given the position of the advocates and their place in global political, economic, and other hierarchies. Its lens is capacity, but also in the normative project of global normative ordering--one that foregrounds risk aversion through the foundational inculcation of the primacy of prevention, mitigation, and remedial strategies that in turn foreground prevention. The essential risk aversion and human centered approach suggests much of what passes for AI regulation, for instance, and is much in evidence in the approaches of the EU (eg here). But it also centers capacity--or relative capacity of human collectives and that, in turn, is understood through the lens of equality that then shapes the discourse of capacity building (and the shifting if its costs).
Sections 1 and 2 of the Report: Artificial Intelligence and the Great Divergence follow below and the entire report may be accessed HERE. And below is its semiotics as imagery.
 |
| Pix credit here |
Council of Economic Advisers 1
1 Introduction
For centuries, most of the world's economies grew at a similarly slow rate. However, a “Great Divergence”occurred with the Industrial Revolution, causing industrializing nations to accelerate their growth relative to the rest of the world.1 Artificial intelligence (AI) is a potentially transformative technology that is often compared to the Industrial Revolution.
However, we are witnessing clear leaders in AI investment, performance, and adoption metrics across
different nations. The Trump administration is laying the groundwork for American AI dominance by
accelerating innovation, infrastructure development, and deregulation while establishing global dominance through technology exports. If the AI revolution is as transformative as the Industrial Revolution, should we expect this to lead to a second Great Divergence? Of course, the future impact of AI is uncertain, so in this paper we focus on the empirical data that can be seen and measured today.
We begin by reviewing analyses of the potential for AI-led economic growth (Section 2) and then discussing estimates of AI’s impact on both GDP and the labor force. Recognizing that these impacts are uncertain and thus need constant monitoring, in Section 3 we highlight metrics for tracking the breakneck pace of investment, performance, and adoption of AI. We then discuss how different countries are faring on these metrics (Section 4). The incredible speed of change cannot be overstated; many of these metrics are doubling every few months and increasing manyfold each year. This means that the AI of the future will likely be very different from the AI of today. We conclude by reviewing the actions President Trump is taking to ensure that America continues to lead on AI (Section 5). As the President said: “America is the country that started the AI race. And as President of the United States, I'm here today to declare that America is going to win it.”2
2 The Future Outlook
The last 25 years have seen a great convergence as the world’s richest nations grew slower than many
developing nations. However, the advent of generative artificial intelligence based around large language
models (LLMs) will initiate a new wave of profound economic transformation in the United States, promising significant boosts to productivity and growth. As AI technologies become more integrated into the workplace, economists are re-evaluating long-term projections for Gross Domestic Product (GDP).
Yet, this period of innovation is not without its complexities. In this report, we focus on the long-term analysis of structural trends, as of course not every AI-related investment will be profitable, and the short-run always contains the potential for substantial volatility.
1 Kenneth Pomeranz, The Great Divergence: China, Europe, and the Making of the Modern World Economy (Princeton University Press, 2000),
https://www.jstor.org/stable/j.ctt7sv80
2 “Trump Advances US Leadership in AI,” Editorials, 2025, https://editorials.voa.gov/a/trump-advances-us-leadership-in-ai/8050987.html
Council of Economic Advisers 2
2.1 Background on Artificial Intelligence
The last few years have seen a rapid explosion in both AI capabilities and jargon, so we begin with a review of several key terms in the AI space.
Artificial intelligence can refer to a wide variety of different computer systems, from chess-playing
computers like Deep Blue to generative AI like ChatGPT. For most of AI’s history, AI was only capable of
making decisions among a relatively small set of options. The recent surge in AI interest has coincided with
the rise of “generative” AI, so called because they are able to “generate” text, images, or video. “Large
language models” are generative AI that can create text.3 They are “large” because of their trillions of
parameters, and “language” because they are trained on large amounts of text written in natural languages.4
5 Agentic AI are a subset of generative AI that go beyond mere content creation and can execute actions in
order to accomplish goals.6
One framework for understanding the intelligence of an AI looks at that intelligence on two dimensions: (1)
its ability to perform different tasks: from writing essays, to identifying objects in pictures, to writing
computer code, to solving math problems and (2) how the AI’s capabilities on that task compare to human-
level intelligence. Today’s artificial intelligence systems have “specialized” (or “narrow”) intelligence
because, although they may be superhuman at a particular task (no human can multiply as fast as a calculator
can), AI is not able to perform all the tasks a human can. Humans are capable of performing a wide variety of
different tasks. Thus, we say that humans have “general” intelligence while current AI (including both
ChatGPT and agentic AI) have “specialized” intelligence.
Artificial general intelligence (AGI) would be a hypothetical AI that can perform all the intellectual tasks that
humans can,7 but the exact definition of AGI is hotly debated, and some definitions only require that AGI
perform “many but not all” human tasks. Artificial superintelligence (ASI), sometimes just called
“superintelligence,” is AI with intelligence that surpasses that of humans.8 The boundary between AGI and
superintelligence is similarly contentious, partly because these terms encompass different aspects of AI:
"AGI" and "specialized AI" describe the generality of tasks an AI can perform, while "superintelligence"
describes the AI’s capabilities on those tasks. However, a “mere” AGI is already superintelligent if it can
perform all human tasks, but at computer speeds. But accounting for semantic disagreements, it is worth
nothing that OpenAI, Anthropic, xAI, Meta, and Google all aim to create artificial general intelligence or
superintelligence.9 10 11 12 13
3 “What is a large language model (LLM)?”, Cloudflare, https://www.cloudflare.com/learning/ai/what-is-large-language-model/
4 “What is a large language model (LLM)?”, Cloudflare, https://www.cloudflare.com/learning/ai/what-is-large-language-model/
5 “What are large language models (LLMs)?”, IBM, https://www.ibm.com/think/topics/large-language-models
6 “What is agentic AI?”, Google Cloud, https://cloud.google.com/discover/what-is-agentic-ai
7 “What is artificial general intelligence?”, Google Cloud, https://cloud.google.com/discover/what-is-artificial-general-intelligence
8 “What is artificial general intelligence?”, Google Cloud, https://cloud.google.com/discover/what-is-artificial-general-intelligence
9 “Planning for AGI and beyond,” OpenAI, February 24, 2023, https://openai.com/index/planning-for-agi-and-beyond/
10 Alex Heath, “Mark Zuckerberg’s new goal is creating artificial general intelligence,” The Verge, January 18, 2024,
https://www.theverge.com/2024/1/18/24042354/mark-zuckerberg-meta-agi-reorg-interview
11 Elon Musk (@elonmusk), “I now think @xAI has a chance of reaching AGI with @Grok 5. Never thought that before,” X, September 17, 2025,
https://x.com/elonmusk/status/1968202372276163029
12 Sarah Perkel, “Anthropic CEO says AGI is a marketing term and the next AI milestone will be like a ‘country of geniuses in a data center,’” Business
Insider, January 22, 2025, https://www.businessinsider.com/anthropic-ceo-calls-agi-marketing-term-2025-1
13 Anca Dragan et al., “Taking a responsible path to AGI,” Google DeepMind, April 2, 2025, https://deepmind.google/discover/blog/taking-a-
responsible-path-to-agi/
Council of Economic Advisers 3
This brings us to an important caveat to this report’s analysis: limitations of economic analysis of artificial
intelligence. As noted by Hanson (2001), artificial intelligence that could perform all human tasks would lead
to absolutely explosive growth and to a very different world than that seen today. Thus, the implications of
AGI (both economic and otherwise) are an important topic deserving of further study, but are generally
outside the scope of our current analysis, which focuses on “narrow” or “specialized” AI.
2.2 Impact of AI on GDP
Economists often think of the productive power of an economy as coming from three factors: the quantity of
labor, the quantity of capital, and total factor productivity (TFP). TFP is a measure of an economy's efficiency
and technological progress. A rising TFP indicates that an economy is producing more goods and services
from the same amount of labor and capital, or the same output with fewer inputs.14 This improvement in
efficiency is a key driver of long-run economic growth and higher living standards.15 For rich countries like
the United States whose capital stocks are already very high, economic growth mainly comes from increasing
total factor productivity.16 17 18
The productivity gains from TFP are eventually translated into higher overall economic output, or GDP.
However, the effect of a new technology occurs with a time lag, as businesses must first successfully adopt
the new technology and adapt their operations.19 Much of the productivity gains in the 1990s emerged from
technological investments that occurred in the 1970s and 1980s.20 Similar technological investments that
occurred during the Great Depression bore fruit during the 1950s and 1960s.21 As a result, while TFP is an
important indicator, it is not a leading indicator of AI’s impact on the U.S. economy. Instead, R&D spending
on AI and the output of AI firms serve as early indicators of technological progress.22 23 For example, AI-
related R&D occurs well before the resulting innovations are widely adopted and have a macroeconomic
effect.
A variety of recent studies have attempted to quantify the impacts of AI on GDP levels. These studies
produced a broad range of estimates: AI could increase GDP by 1 percent up to more than 45 percent. The
wide range reflects the high degree of uncertainty surrounding the economic characteristics of AI. However,
14 Robert Zymek, “Total Factor Productivity,” IMF, September 2024, https://www.imf.org/en/Publications/fandd/issues/2024/09/back-to-basics-
total-factor-productivity-robert-zymek
15 Robert Zymek, “Total Factor Productivity,” IMF, September 2024, https://www.imf.org/en/Publications/fandd/issues/2024/09/back-to-basics-
total-factor-productivity-robert-zymek
16 Robert Shackleton, “Total Factor Productivity Growth in Historical Perspective,” Congressional Budget Office, March 2013,
https://www.cbo.gov/sites/default/files/113th-congress-2013-2014/workingpaper/44002_TFP_Growth_03-18-2013_1.pdf
17 Edward C. Prescott, “Needed: A Theory of Total Factor Productivity,” International Economic Review, August 1998,
https://www.jstor.org/stable/2527389
18 Scott A. Wolla, “What Are the ‘Ingredients’ for Economic Growth?”, Federal Reserve Bank of St. Louis, September 1, 2013,
https://www.stlouisfed.org/publications/page-one-economics/2013/09/01/what-are-the-ingredients-for-economic-growth
19 Wenjie Tang, Tong Wang, and Wenxin Xu, “Sooner or Later? The Role of Adoption Timing in New Technology Introduction.” Production and
Operations Management, April 2022, https://onlinelibrary.wiley.com/doi/epdf/10.1111/poms.13637?msockid=28439e724fd560f012f588f14e1861b7
20 Roger W. Ferguson Jr. and William L. Wascher, “Distinguished Lecture on Economics in Government: Lessons from Past Productivity Booms,”
Journal of Economic Perspectives, 2004, https://www.federalreserve.gov/boarddocs/speeches/2004/20040707/attachment.pdf
21 Roger W. Ferguson Jr. and William L. Wascher, “Distinguished Lecture on Economics in Government: Lessons from Past Productivity Booms,”
Journal of Economic Perspectives, 2004, https://www.federalreserve.gov/boarddocs/speeches/2004/20040707/attachment.pdf
22 Luisa R. Blanco, Ji Gu, and James E. Prieger. “The Impact of Research and Development on Economic Growth and Productivity in the U.S. States,”
Southern Economic Journal, January 2016, https://onlinelibrary.wiley.com/doi/abs/10.1002/soej.12107
23 Yen-Chun Chou, Howard Hao-Chun Chuang, and Benjamin B.M. Shao. “The Impacts of Information Technology on Total Factor Productivity: A
Look at Externalities and Innovations,” International Journal of Production Economics, December 2014,
https://www.sciencedirect.com/science/article/abs/pii/S0925527314002618
Council of Economic Advisers 4
it is worth noting that in the first half of 2025 alone, AI-related investment increased GDP by an annualized
rate of 1.3 percent, harkening back to the scale of railroad investment during the Industrial Revolution24 25 and
seemingly ruling out the lowest few estimates. Mid-range estimates for the effects of AI on GDP include
those from a variety of companies such as Oxford Economics (1.8 to 4 percent increase after 8 years),
McKinsey (2.4 to 4.1 percent increase in the long run) and Goldman Sachs (7 percent increase after 10 years).
High estimates include those by PricewaterhouseCoopers (8 to 15 percent after 10 years) and a BIS Academic
Working Paper by Aldasoro et al. (20 to 45 percent after 10 years for their approaches that assume all sectors
of the economy will be at least somewhat impacted by AI). Alonso et al. have a wide range of estimates (4.7
to 19.5 percent), reflecting uncertainty over whether AI will substitute more for skilled or unskilled labor (the
latter of which would yield the divergence and therefore the high-end growth estimate for the U.S.). For
comparison, a 2010 ITIF study indicated that the IT revolution boosted U.S. GDP by about 14 percent.26 27
These estimates all assume that AI can partially but not completely substitute for human labor: in the case
where AI could do all human tasks, capital becomes a substitute for labor and economic growth increases to
45 percent per year (see Hanson, 2001).
24 Page 5 of Rui M. Pereria et al., “Railroads and Economic Growth in the Antebellum United States,” College of William and Mary Department of
Economics, December 2014, https://economics.wm.edu/wp/cwm_wp153.pdf
25 Note that this 1.3 percent value for AI is the impact of AI investment on the level of GDP, even before any productivity gains from that investment
are reaped. CEA staff could not locate this exact statistic for railroads during the Industrial Revolution, but U.S. investment in railroads roads grew
from 0.2 percent of GDP in 1830, to 0.9 percent in 1839, to a maximum of 2.6 percent of GDP in 1854 (Pereria et al., 2024).
26 Robert D. Atkinson et al., “The Internet Economy 25 Years After .Com,” The Information Technology and Innovation Foundation, March 2010,
https://www2.itif.org/2010-25-years.pdf
27 ITIF indicates $2 trillion. The report was published in early 2010. 2009 U.S. GDP was $14.5 trillion, see “Gross Domest
No comments:
Post a Comment