# Samo Burja on Growth, Energy, and AI

Podcast: The a16z Show
Published: Jun 17, 2026
Reading time: 15 min
Canonical: https://podbrew.app/briefs/the-a16z-show-samo-burja-on-growth-energy-and-ai

Theo Jaffee welcomes Samo Burja, founder of Bismarck Analysis, to discuss the profound implications of AI. They delve into how AI's escalating demands for compute power, energy, and infrastructure are poised to reshape global industrial capacity and economic growth.

The discussion centers on AI as more than a software story; it is a significant demand shock that will ripple through physical sectors. This includes energy, manufacturing, construction, and global supply chains, potentially triggering a new wave of industrial expansion far beyond traditional technology industries.

Understanding these forces is critical as Burja shares his perspectives on functional institutions, state capacity, and the political economy of automation. Their conversation highlights why societies that can effectively organize people and resources for this new era may gain a lasting advantage.

## Key takeaways

- The sheer scale of AI's demands is prompting industrial revolutions in fundamental industries like silicon, steel, and natural gas.

- The primary impact of AI is shifting from purely software-centric to a physical one, requiring massive increases in energy, compute, and industrial capacity.

- Societal beliefs are more influenced by concrete material conditions than solely by deep-seated religious traditions.

- As society secularized, capitalism reinforced and prolonged this hard work ethic through compounding material rewards.

- High costs in cities with older, limited housing stock, such as San Francisco, lead to inflated prices for everyday services, like $20 lattes, creating economic pressure for change.

- The rising cost of human labor in expensive regions makes automation an increasingly economically rational solution, pushing towards the adoption of robotics in various industries.

- Investing these newly printed funds into AI companies via equity purchases could be a more effective economic stimulus and industrial policy compared to past initiatives.

- A booming economy, potentially spurred by investments in frontier AI, is presented as the most crucial asset for ensuring the long-term sustainability of a welfare state.

- Political incentives often hinder the implementation of long-term economic strategies, as politicians typically prioritize short-term gains over distant benefits.

- A population's political power is tied to its utility as conscripts, taxpayers, or its potential to riot and overthrow the government.

- The historical automation of agriculture in the US profoundly reshaped the rural economy and drove mass migration to urban centers.

- The composition of a country's labor force dictates who is best compensated and most politically useful, constantly reshaping national politics even if constitutional frameworks remain unchanged.

- Industrial economies like Taiwan and South Korea are poised for significant economic growth, contrasting sharply with knowledge economies that face increasing concerns about mass unemployment.

- Global white-collar professions, from finance in London to law in Geneva, face significant disruption as frontier AI models allow top firms to scale services dramatically.

- The economic impact of AI will create a "winner-take-all" environment in white-collar sectors, where a few highly capable, AI-powered firms dominate, potentially reducing the need for many human professionals.

- Contrary to white-collar disruption, industrial economies like Taiwan and South Korea are poised for 10% growth due to high demand for memory chip production, linking to the expansion of companies like ASML.

- AI's demand for advanced technology can significantly boost the economies of key suppliers like ASML's home country, the Netherlands.

- Critically low fertility rates in major industrial economies like Taiwan (0.65) and South Korea present a significant demographic headwind that will eventually curb their sustained domestic growth.

- Implementing AI into broken processes will only amplify existing organizational bottlenecks.

- Functional institutions are best equipped to adjust to and benefit from AI effectively.

## 00:05 - 02:05 AI's Massive Demands Are Fueling a New Industrial Revolution

AI's enormous computational needs are so vast that they are driving demand for industrial revolutions across various sectors beyond just software. This represents a significant macroeconomic shift.

This 'demand shock' originating from AI is initially impacting silicon but will eventually extend to foundational materials like mirrors, steel, and natural gas. This widespread demand signifies a rekindling of industrial revolutionary processes.

While most discussions about AI focus on software, the larger implication lies in its physical demands for energy, compute, data centers, and chip production. This growth will reshape manufacturing, construction, energy, and the broader global economy.

> The demands of AI are so massive that for the first time in decades, the economies of scale necessary to supply them require industrial revolutions in everything.

## 02:05 - 04:05 Concrete material conditions, not just religious traditions, shape societal beliefs over time.

The assumption that deep religious traditions solely dictate a society's core beliefs understates the impact of concrete material conditions. Religious landscapes are dynamic, changing significantly over centuries or even decades, which often goes unrecognized.

For example, the 3rd-century Yellow Turban Rebellion in China was a Daoist cult that believed in an end-of-the-world scenario and sought to transform society into an unprecedented utopia. This movement led to a bloody civil war, likely destroying the Han dynasty, demonstrating that eschatological thinking is not exclusive to specific religious or societal structures.

In the United States, the religious landscape has evolved considerably, even in the last fifty years. Many nominally Christian individuals now hold a non-denominational, Christian-derived ethic emphasizing being a "good person," coupled with a vague belief in karma and an inactive God who does not interfere in the world. This practical religious sense continues to innovate within the country.

> I think that people want to equate deep religious traditions with Basically the beliefs that a society holds, but they underrate how much those religious beliefs change on the century timescale or the millennial timescale.

## 04:05 - 06:05 Puritanism's Enduring Influence on American Capitalism

Early American Puritanism, rooted in Calvinist beliefs, profoundly influenced the nation's economic landscape. A key tenet was predestination, the idea that individuals were born already chosen for heaven or hell. While material success didn't guarantee salvation, it was often interpreted as a hint or a possible sign that one belonged to the "elect."

This theological framework gave rise to the Protestant work ethic, which viewed diligent labor in one's vocation as a divine calling. The pursuit of material well-being, therefore, became intertwined with spiritual assurance, motivating early settlers to work exceptionally hard and succeed in their chosen professions.

As American society gradually secularized, the initial religious motivation for hard work diminished, but the habit persisted. The tangible material rewards generated by this intense work ethic began to compound, reinforcing the behavior. Capitalism, initially nurtured by these values, then took on a life of its own, perpetuating and developing the very work ethic it inherited.

Today, the legacy of this historical dynamic means many continue to work hard under a general assumption of future success, connecting to the long-standing "land of opportunity" narrative, even if its present-day reality is debated.

> I think today people work hard on the vague assumption that, you know, it's going to work out for you, right?

## 06:05 - 08:05 China's Economic Boom and the Rise of Automation

China's recent economic trajectory for most of its citizens is comparable to 1950s America, where living standards have consistently improved over the past 45 years. Families are acquiring their first cars, and urban living space has increased, surpassing some lower-ranking European countries. This rapid development contrasts with the generally older housing stock found in many established Western cities.

While China has seen extensive new construction, accommodating its growing and increasingly affluent population, cities like San Francisco, by contrast, feature much older housing, with many historic Victorians dating back to 1906. This limited and aging housing stock contributes to exceptionally high living costs in such areas.

The soaring cost of living in places like San Francisco drives up the price of services, making items like lattes increasingly expensive. This economic pressure makes automation an increasingly rational and attractive solution. As human labor costs rise, robots become a more economically viable alternative for various tasks, indicating a potential shift towards greater robotic integration in high-cost environments.

> If you make housing very expensive, then you're gonna see fifty-dollar lattes eventually. I mean, you're already seeing twenty-dollar lattes in the city. I think fifty isn't far away, so, you know, eventually the robot makes economic sense.

## 08:05 - 10:06 US Government Investment in AI Companies as Economic Stimulus

The US government has a history of injecting trillions of dollars into the economy, often in response to crises like COVID or to support the stock market and 401k holders. This trend of printing large sums of money, driven by political incentives to prevent market downturns, is expected to continue.

Samo Burja suggests that instead of traditional stimulus packages or infrastructure bills, directing these trillions towards AI companies could be a superior approach. He proposes that the government could print a trillion dollars and use it to purchase equity in leading AI firms such as OpenAI or Anthropic.

This form of investment would serve as both an economic stimulus and industrial policy. The funds would likely be channeled into critical areas like employee compensation and the development of data centers, which are fundamental to AI advancement. This directly contributes to technological progress and economic activity, potentially yielding better returns than previous spending initiatives.

> printing a trillion dollars and paying it to AI companies in exchange for equity might be the best economic stimulus the US government can carry out and might yet prove a vastly better investment than, say, the infrastructure bill was.

## 10:06 - 12:06 Government AI Investment as a Long-Term Strategy for Welfare

The discussion contrasts direct government welfare payments with strategic investments in frontier AI labs, positing that the latter could be a more effective long-term approach to sustaining a welfare state. The argument emphasizes that a thriving economy is the most valuable asset for a welfare system's endurance, rather than relying solely on immediate handouts.

Countries like Sweden and Denmark are cited as examples of nations that historically understood this principle, developing export-driven economies that fuel their welfare programs. Sweden, for instance, is noted for its capitalism and global companies like IKEA, which underpin its social services, demonstrating that a robust economic engine can coexist with and support extensive welfare provisions.

While such long-term investments in AI infrastructure are seen as beneficial, the challenge of political incentives is acknowledged. Politicians typically favor short-term decisions, which may not align with the delayed gratification of large-scale technological investment. However, a cynical perspective suggests that corporate handouts to profitable companies, driven by a "bipartisan corruption consensus," might unintentionally foster economic growth that could ultimately benefit the welfare state more than direct dole payments.

This perspective highlights a tension between immediate social support and the foundational economic growth necessary for a truly sustainable welfare system, suggesting that current tax systems, by punishing profitability, might even work against this long-term view.

> The best asset for a welfare state, if you want it to be sustainable, is a booming economy.

## 12:06 - 14:06 OpenAI's Strategic Proposal for Government Ownership

OpenAI's proposal for partial government ownership is a calculated move designed to enhance its public image and outflank competitors. While companies like Anthropic have attempted to position themselves as the 'least disfavored' AI company by acknowledging concerns about inequality, OpenAI's direct offer of government ownership takes this strategy further, portraying the company as more benevolent to the general public.

This proposal transforms OpenAI, an American frontier lab, into a system resembling a state-owned enterprise. The current US law suggests the government might have to buy out the company if such a proposal were accepted, potentially leading to a massive government investment.

Such a move could be a highly effective business decision for OpenAI, positioning it as the most favored government provider and leveraging a national security argument. While beneficial for the company, its implications for the broader landscape of capitalism or the trajectory of technological progress are debated, particularly when considering the future impact of full labor automation on the human workforce and societal structures.

> I actually think that the transformation of an American frontier lab into this weird state-owned enterprise type of system might actually be a great business decision for the company.

## 14:06 - 16:06 Automation's Historical Impact on Political Power and Labor

Samo Burja proposes that a populace's political power is fundamentally linked to its utility to the state, whether through military service, tax contributions, or the capacity to instigate social unrest. He suggests that the power of the "welfare voting class" is tied to its ability to organize protests.

He contends that even partial automation has a profound effect on a nation's political economy, citing historical shifts in the US as evidence. Burja specifically refers to the New Deal era as a period where agricultural automation significantly transformed the country.

The automation of agriculture made many small, "mom and pop" farms economically unviable, leading to a substantial migration of the population from rural areas to cities. This shift dramatically altered the composition of the labor force and, consequently, the distribution of political influence within the nation.

> I just don't understand why the voters would keep control of a state, right? I think people have control of states because they're either useful as conscripts, useful as taxpayers, or potentially because they can riot in the street and, you know, overthrow the government.

## 16:06 - 18:07 Industrial Economies Project Growth While Knowledge Economies Face Unemployment Risks

Globalization, particularly the rise of East Asian industrial complexes like China, significantly outcompeted American blue-collar workers. This transformation led to the weakening of industrial unions, making political promises for a revival of the American unionized worker class seem anachronistic.

The economic landscape is shifting dramatically, with different outcomes for industrial versus knowledge economies. Industrial economies, exemplified by countries like Taiwan, South Korea, and the Netherlands (with ASML), are projected to achieve substantial 10% year-on-year economic growth.

Conversely, nations relying on a knowledge economy model are increasingly concerned with the prospect of mass unemployment. This divergence raises questions about how these economic shifts will interact with declining fertility rates, which could potentially lead to permanent recessions.

> If your economy is a knowledge economy, you're worrying about mass unemployment. In contrast, if your country has an industrial economy, you're looking at ten percent year-on-year economic growth.

## 18:07 - 20:08 AI disruption targets global white-collar jobs, simultaneously fueling industrial economic growth.

Frontier AI models are poised to significantly disrupt white-collar professions worldwide. Individuals working in finance in London or law in Geneva, whose companies lack access to these advanced AI capabilities, are likely to face considerable challenges. This disruption suggests a shift in the global employment landscape for these sectors.

The impact will create a winner-take-all scenario, particularly for top-tier firms in major hubs like San Francisco or New York. For example, if a prestigious American law firm can leverage AI to serve a million customers instead of just five hundred or five thousand, there would be little reason to choose any other firm. This model suggests that AI could handle 99% of client interactions, with human lawyers providing brief oversight.

However, this white-collar upheaval is contrasted by an expected boom in industrial economies. Countries like Taiwan and South Korea are projected to achieve 10% economic growth in the short term, driven largely by high demand for memory chip production. This growth is directly linked to the expansion of key industrial players, such as ASML, illustrating a divergent economic trend fueled by technological advancement.

> If America's most prestigious law firm can handle a million customers rather than five hundred or five thousand customers, why would you ever hire anyone else?

## 20:07 - 22:08 AI's demand shock can revive industrial economies

ASML, a leading chip equipment manufacturer with a market cap around $650 billion, could drive significant economic growth in the Netherlands, potentially up to 10%. This could provide a much-needed cash infusion to address challenges like an aging population. The company also fosters a vast industrial ecosystem, relying on many other Dutch and German companies, especially in the optics sector.

This ripple effect extends to Germany, which, despite its larger economy and high energy costs, could see an additional three or four points of economic growth due to AI-driven demand. The immense demands of artificial intelligence are creating such a profound demand shock that industrial revolutions are becoming necessary across various sectors to meet the required economies of scale.

The underlying macro story suggests that AI's insatiable need for resources and infrastructure is so massive it can rescue and bolster even declining industrial bases. This represents a significant shift, as for the first time in decades, the scale of demand necessitates widespread industrial revitalization.

> the demands of AI are so massive that for the first time in decades, the economies of scale necessary to supply them require industrial revolutions in everything.

## 22:08 - 24:08 Low Fertility Rates Pose a Major Headwind to Industrial Growth

The current surge in demand, partly fueled by AI, is driving a new wave of industrial activity, requiring a significant build-out of resources like steel, cement, and construction labor. This effectively reignites aspects of the industrial revolution.

However, the long-term sustainability of this industrial growth faces a critical demographic challenge: severely aging populations and extremely low fertility rates, particularly in leading industrial nations. This acts as a significant headwind, slowing progress.

Countries like South Korea and Taiwan, despite their current economic prowess, are projected to experience a halt in their domestic growth due to these demographic trends. Taiwan, for instance, now has one of the lowest fertility rates globally at 0.65, even lower than Korea.

To maintain industrial output, these nations will increasingly need to offshore and outsource production to countries like Indonesia or India, where populations are younger and growing, shifting economic opportunities elsewhere.

> Taiwan, of course, in contrast with their, you know, ten percent GDP growth rate, has the lowest fertility rate In the world, it's like point six five now, it's lower than Korea.

## 24:08 - 26:09 Functional institutions are crucial for leveraging AI effectively.

AI provides the ability to scale white-collar labor, presenting a significant opportunity for organizations to enhance productivity and efficiency across various sectors.

However, simply integrating AI into a system that already has broken processes will not resolve underlying issues. Instead, it will act as a magnifying glass, exposing and exacerbating existing bottlenecks within a company or organization.

Therefore, the entities that will most successfully navigate and benefit from the AI era are those with robust, functional institutions. These organizations, whether private companies, government bureaucracies, or other groups, are better positioned to adapt their operations and integrate AI seamlessly.

> If you have broken processes and you try to fit AI into a broken process, you are just putting the weight into all the remaining bottlenecks inside your company or organization.

---

Get podcast briefs for shows you follow: https://podbrew.app/