AI Is Ushering in an Entirely New Economic Paradigm | Jordi Visser

Key Concepts

• Infinite Need for AI: The speaker posits that the demand for Artificial Intelligence is boundless, akin to solving all problems in existence.
• Intelligence Plugging: AI is being integrated into all existing machines to enhance their efficiency.
• Deflationary Spiral: Increased intelligence growth is seen as a precursor to a deflationary spiral.
• Two-Part Economy: The economy is divided into the digital economy and the traditional industrial economy, with the digital sector growing in significance.
• Profit Margins as Measurement: Profit margins are considered a more accurate metric for measuring productivity gains than GDP.
• Visual Language Models (VLMs): The next evolution of AI beyond Large Language Models (LLMs), incorporating visual understanding.
• AI as an Endgame: AI is viewed as a tool for solving the world's most significant problems.
• Government Intervention in AI: The US government's executive order on AI signifies a critical public-private partnership, making AI development a national security priority.
• Deflationary Pressures: AI's impact on job markets and efficiency is expected to lead to deflationary pressures.
• Bitcoin as a Store of Value: Bitcoin is presented as a digital store of value, analogous to gold, representing the digital economy.
• Democratizing Force of Technology: Technology, including AI and DeFi, is seen as empowering individuals in developing nations.

AI and the Infinite Need for Intelligence

The core argument presented is that the demand for Artificial Intelligence is effectively infinite. This is not a finite problem to be solved but rather an ongoing process of enhancing efficiency across all machines ever built. As AI systems become more intelligent, they will uncover solutions that are currently unimaginable. This continuous growth in intelligence is posited to accelerate the economy towards a deflationary spiral. The speaker emphasizes that this trend is unstoppable and viewing it as a bubble is a mistake.

The Evolution of Economic Measurement: Beyond GDP

The traditional global macro frameworks, often reliant on indicators like the yield curve, have struggled to accurately predict economic outcomes in recent years. This is attributed to a fundamental shift in the economy, characterized by the rise of the digital economy. The speaker highlights that GDP, as conceptualized by Simon Kuznets, was not designed to measure intangibles like software and digital services. The proliferation of smartphones and apps has replaced numerous physical goods, creating a two-part economy where the digital sector's contribution is increasingly significant but poorly captured by old metrics.

A key example is the prolonged negative LEI (Leading Economic Index) without a corresponding recession, which the speaker attributes to the difficulty in measuring the digital economy's impact. Furthermore, the speaker argues that household net worth, now significantly larger than the economy's GDP, provides a buffer that traditional recession indicators fail to account for. Transfer payments are also noted as a growing component of income, further complicating traditional economic analysis.

Profit Margins as a Superior Metric

Instead of GDP, the speaker advocates for using profit margins as a more accurate measure of economic productivity and growth. This is because profit margins have a tangible accounting basis and have been consistently rising for the past 40 years, a trend that began with the advent of coding in the early 1980s. The speaker points to the stark contrast in profit margins between the non-coding sectors of the economy (low) and high-tech companies like Nvidia and the Mag 7 (enormous), and even private companies like Stripe and Cursor, which exhibit almost infinite profit margins due to their lean operational models.

The AI Buildout: From LLMs to VLMs and Beyond

The current AI buildout, particularly the significant capital expenditure on data centers, is framed not as a cost but as the creation of "gigantic brains" that will solve the world's greatest problems. The evolution of AI is moving beyond Large Language Models (LLMs) to Visual Language Models (VLMs), which integrate textual understanding with visual intuition. This transition requires significantly more computational power (50 to 1000 times more) to process and learn from the real world. The ultimate goal includes visual language action for humanoids and military applications, necessitating on-board intelligence that doesn't rely on external networks.

Infinite Revenue Streams in AI

The speaker identifies drug discovery and healthcare as areas with potentially infinite revenue streams driven by AI. The ability to cure diseases like cancer and address aging is seen as a massive economic opportunity. The speaker contrasts the current AI buildout with the dot-com bubble, noting that the former is funded by AI experts using their accumulated capital, unlike the telecom-funded buildout of the dot-com era. The race for AI supremacy is driven by military objectives, legacy building, and solving fundamental problems like cancer.

The Transition to Credit Markets and Profit Margin Compression

While companies have largely self-funded the current AI buildout through substantial free cash flow, a transition towards tapping credit markets is anticipated. The speaker predicts a compression of profit margins across the board, as not all companies will justify their spending. The risk associated with AI spending is not immediate but will grow over time. As AI intelligence advances exponentially, it will lead to solutions that reduce the need for certain types of infrastructure, potentially leading to a deflationary spiral. This is distinct from an asset crisis, as profit margins for public companies may continue to grow, but the number of "losers" in the market will accelerate.

Nvidia: A Moat in the Data Center Buildout

Nvidia is presented as a company with a significant moat in the data center buildout, holding approximately a 90% market share. Despite the emergence of competitors like Google's TPUs, Nvidia's established experience and dominance in massive data center clusters make it the preferred choice for companies racing to build AI infrastructure. Even with a projected decrease in market share to 50%, Nvidia's revenue projections remain exceptionally strong, making it appear "dirt cheap" relative to its long-term potential.

The Dot-Com Bubble Analogy and Infinite Demand

The speaker revisits the dot-com bubble analogy, arguing that while the buildout of infrastructure (like dark fiber) was ahead of demand, it was eventually utilized. In contrast, the demand for AI is considered truly infinite, akin to the need for electricity. Capacity constraints are already appearing across cloud providers, indicating a persistent and growing demand. The transition from LLMs to VLMs further amplifies this demand for compute.

Government Intervention and the Genesis Mission

A significant development highlighted is the US government's executive order on AI, termed the "Genesis Mission." This order signifies a merging of public and private sectors, making AI development a critical national security priority, especially in competition with China. The initiative involves opening national labs, data, and supercomputers to private sector foundation models, and it is expected to address power constraints and accelerate AI development across various fields, including biology, science, and advanced materials. This government backing is seen as making AI development "untouchable" and a key driver for national survival.

Deflationary Risks and the Future of Monetary Policy

The speaker identifies deflation as a significant risk over the next few years, driven by AI's impact on job markets and efficiency. This contrasts with the current focus on inflation and debt sustainability. The traditional tools of monetary policy are seen as increasingly inadequate in this new paradigm. The speaker argues that the government will likely find ways to implement forms of Universal Basic Income (UBI) and that the focus should shift from short-term economic data to understanding these larger, transformative trends.

The Utility Sector and Inflation Concerns

Concerns about inflation driven by the utility sector are dismissed. The speaker argues that hyperscalers will absorb any cost increases, and electricity is a minor component of overall costs for AI companies. Furthermore, utilities are not a major driver of CPI, and with falling housing costs, stable gas prices, and potentially declining wages, the argument for significant inflation is weak.

The Role of Technology in Addressing Societal Issues

The speaker touches upon Michael Green's analysis of the poverty line, highlighting how factors like childcare, demographics, wealth inequality, obesity, and immigration contribute to societal challenges. AI's potential to address these issues, coupled with deflationary pressures from job displacement, creates a complex socio-economic landscape. The speaker suggests that a significant portion of the population may become "trapped" in a world of limited economic opportunity, contributing to the rise of socialist sentiments.

Debt Sustainability and the Longevity Revolution

The speaker argues that concerns about debt sustainability are overstated due to the potential for a "longevity revolution" driven by AI. The ability to cure diseases and reverse aging could lead to people living indefinitely, fundamentally altering economic and demographic assumptions. This, combined with deflationary pressures, could alleviate debt burdens. The speaker believes that government spending has been largely directed towards supporting aging populations and maintaining votes, a trend that could be disrupted by increased lifespans.

Bitcoin: The Digital Store of Value and the Fastest Horse

Bitcoin is presented as the only accepted store of value in the digital economy, analogous to gold's historical role. Its scarcity and resilience through time have established it as a significant asset. The speaker draws a parallel between the government's debasement of fiat currency and the "zombification" of businesses by dominant tech companies, leading to Michael Saylor's investment in Bitcoin as a hedge. Bitcoin's scarcity is seen as a crucial differentiator in a world of increasing money printing.

The Democratizing Force of Bitcoin and DeFi

The speaker believes that Bitcoin and decentralized finance (DeFi) are democratizing forces, providing access to financial services and entrepreneurial opportunities for individuals in developing nations who distrust their governments and are excluded from traditional fiat systems. This trend is expected to continue for the next 20 years, with significant capital migrating into Bitcoin.

The Financialized Economy and the "Video Game" Market

The speaker posits that the economy has become highly financialized, resembling a "video game" where governments intervene to prevent significant downturns. This intervention, exemplified by the Fed's willingness to buy corporate bonds, means that traditional crisis scenarios are unlikely. Younger generations, accustomed to digital interactions, are more accepting of this financialized environment and assets like Bitcoin, while older generations may struggle to reconcile it with traditional valuation methods. The speaker concludes that the market is not fair but rather managed to prevent collapse, with Wall Street acting as the "tip of the spear" that eventually supports Main Street.

Bitcoin's IPO Moment and Distribution

The "IPO moment" for Bitcoin is discussed in the context of its increasing acceptance and the need for wider distribution. The speaker believes that a concentration of ownership among a few individuals is a weakness and that a more equitable distribution is necessary for Bitcoin to mature as an asset. The potential for the poorest individuals globally to invest in Bitcoin and participate in the digital economy is highlighted as a key aspect of its democratizing influence.