AI Apocalypse, a massive short squeeze

Original Title: It's Too Obvious. What If AI Doesn't Actually End The World?

Original Author: The Kobeissi Letter

Original Translation: DeepTech TechFlow

DeepTech Summary: As AI tools like Anthropic demonstrate astonishing capabilities in code and workflow automation, the market has plunged into "AI Doomsday" panic, causing the evaporation of trillions of dollars in market value. However, this article presents a highly insightful reverse perspective: the short-term impact triggered by AI is not a harbinger of economic collapse but a significant reduction in the "cognitive cost." By contrasting the 1980s PC revolution and productivity historical data, the author points out that when technology makes knowledge acquisition cheap and abundant, the true era of "abundant GDP" will begin. This is not just a labor force reconfiguration but also the path to geopolitical easing and a global productivity boom.

Full Text Below:

The stock market just wiped out $800 billion in market value as the "AI taking over the world" narrative is becoming a consensus view. This view is too obvious. And trades based on the obvious never truly win.

The reason this doomsday scenario has gone viral is that it captures something instinctual. It portrays AI not as a productivity tool but as a macroeconomic destabilizer capable of triggering a negative feedback loop: layoffs lead to weakened consumption, weakened consumption leads to more automation, and automation further accelerates layoffs.

The obvious fact is: AI is not just another software feature or efficiency enhancement tool. It is a universal capability shock that simultaneously touches every white-collar workflow. Unlike any revolution in history, AI is becoming good at "everything."

But what if the doomsday scenario is wrong? It assumes demand is fixed, productivity gains do not expand the market, and the rate of system adaptation cannot exceed the rate of destruction.

We believe there is a second path, one that is vastly underestimated. The Anthropic "takedowns" that look like early signs of systemic collapse may ultimately be the beginning of the largest-scale productivity expansion in history.

Before we begin, bookmark this article and revisit it repeatedly over the next 12 months. While the analysis below is not a definite outcome, it is crucial to remember that humans always find a way to turn the tables, and free markets always have a self-healing capability.

Anthropic's "Unpacking" Is Real

First, we must acknowledge that we cannot ignore the market. Anthropic is disrupting the world through Claude, causing Fortune 500 companies to lose hundreds of billions in market cap.

This is a story we've seen several times in 2026: Anthropic releases a new AI tool, Claude makes substantial advancements in programming and workflow automation, and within hours, the market of the target industry collapses.

If you have not been paying attention, here are some examples:

Stock Reactions to Claude's Announcements

IBM stock ($IBM) just had its worst day since October 2000 after Anthropic announced that Claude could streamline COBOL code. Adobe ($ADBE) is down -30% year-to-date as generative capabilities compress creative workflows. The cybersecurity sector crashed after the release of "Claude Code Security."

In the above example, CrowdStrike stock ($CRWD) plummeted almost instantaneously when Claude announced "Claude Code Security."

On February 20 at 1:00 p.m. ET, Claude announced "Claude Code Security." This is an automated AI tool that can scan code repositories for vulnerabilities.

Just two trading days later, CrowdStrike stock ($CRWD) had evaporated -20 billion in market cap due to the impact of this news.

These reactions are not irrational. The market is attempting to price in real-time profit compression. When AI replicates worker tasks, pricing power shifts to the buyer. This is the first-order effect, and it is very real.

Commoditization does not equal collapse. Instead, it is a way to lower costs and broaden access through technology. Personal computers commoditized computation, the internet commoditized distribution, the cloud commoditized infrastructure, and AI is now commoditizing cognition.

Undoubtedly, some traditional workflows will experience margin compression. The question is, will the lower cognitive cost lead to an economic collapse, or will it be allowed to expand dramatically?

The "Doom Loop" Assumes Demand is Fixed

The doomsayer's loop has created a simplified linear model: AI improves, companies reduce layoffs and wages, then purchasing power declines, companies reinvest in AI to defend profits, and so the cycle goes on. This assumes a completely stagnant economy.

History shows that this is not the case. When the cost of producing something collapses, demand rarely remains the same but expands instead. When the cost of computation decreases, we don't consume the same amount of computation at a cheaper price. We consume orders of magnitude more computation and build entirely new industries on that basis.

As shown in the chart below, the price of a personal computer today is 99.9% cheaper than in 1980.

Chart Note: 1980-2015 Personal Computer Price Trend

AI reduces the cost across every industry, and when the cost of services decreases, purchasing power increases regardless of whether wages grow.

The doomsday loop only dominates when AI replaces labor without substantial expanded demand. If cheap computation and productivity give rise to entirely new categories of consumption and economic activity, then an optimistic scenario emerges.

The Real Impact is Price Collapse, Not Unemployment

Investors find it easier to sell the "obvious" layoff story, but the price compression that the service sector is undergoing is the bigger news. Knowledge work is expensive primarily because of the scarcity of knowledge—that sounds trivial but it is the actual case. The abundance of knowledge supply has led to a decline in the price of knowledge work.

Think of medical management, legal documentation, tax filing, compliance checks, marketing production, basic coding, customer service, and education tutoring. These services consume a significant amount of economic resources largely because they require trained human attention. AI has lowered the marginal cost of this attention.

In fact, as shown in the chart below, the U.S. service sector contributes to nearly 80% of the U.S. GDP.

If the cost of running a business decreases, small businesses become more accessible; if the cost of accessing services decreases, more households partake. In a way, the progress of AI can act as a form of "invisible" tax cut.

Companies whose profits rely on high-cost cognitive labor may suffer losses, but a broader economy will benefit from lower service inflation and higher real purchasing power.

Shifting from "Ghost GDP" to "Abundance GDP"

The bear case hinges on "Ghost GDP," output that shows up in the data but does not benefit households. The optimistic rebuttal is what we call "Abundance GDP," where output growth combines with falling living costs.

"Abundance GDP" doesn't require nominal income spikes; it requires prices to fall faster than incomes. If AI lowers the cost of many essential services, then even as household wage growth slows, their real income rises. Thus, the productivity gains do not disappear; they are passed on through lower prices.

This may explain why productivity has outpaced wage growth over the past 70+ years:

The internet, electricity, mass manufacturing, and antibiotics have all provided new ways to expand output and reduce costs, despite these processes being disruptive and volatile. However, in hindsight, these changes permanently elevated living standards.

A society that reduces time wasted on navigating complex systems and redundant services in payments becomes functionally richer.

The Labor Market Is Restructuring, Not Disappearing

A core concern is that AI will disproportionately impact white-collar employment, which drives nonessential consumption and housing demand. This is a fact and a valid concern, especially against the backdrop of such significant wealth inequality.

However, AI faces more challenges in the physical world's dexterity and human identity. Skilled trades, hands-on healthcare, advanced manufacturing, and experience-driven industries still maintain structural demand. In many cases, AI is a complement to these roles, not a replacement.

More importantly, AI lowers the barrier to entry for entrepreneurship. When one can automate accounting, marketing, support, and programming tasks, starting small-scale businesses becomes easier. We are bullish on small businesses.

In fact, by eliminating entry barriers through AI, we may be solving the wealth gap issue we currently face.

The internet killed certain job categories but created entirely new ones. AI may follow a similar pattern, compressing some white-collar functions while expanding self-directed economic participation in other areas.

Acknowledged, proceeding to modularize and compile Part Three (final part) for you. This section will explore the evolution of the SaaS business model, AI's reshaping of market structures, real-world productivity data, and an underestimated perspective: how AI-driven "abundance" can reduce global conflicts.

The "Demise" Story of SaaS

AI has clearly put pressure on the traditional SaaS (Software as a Service) business model. Procurement team negotiations have become more challenging, and some niche software products face structural resistance. However, SaaS is just a delivery mechanism, not the endpoint of value creation.

The next generation of software will be adaptive, agent-driven, outcome-based, and deeply integrated. The winners will be not just providers of static tools but those most adaptable to change.

Every technological shift rearranges the stack, and companies pricing for static workflows will struggle. Those with data, trust, compute, energy, and validation may thrive.

A compression of profits at one layer does not mean the collapse of the entire digital economy; it signifies transformation.

AI Business Repositioning Markets

Doubters believe that Agentic Commerce will destroy middlemen and eliminate fees. To some extent, it does. When friction decreases, extracting fees becomes more challenging.

As shown in the chart below, even before AI became what it is today, stablecoin transaction volumes were already skyrocketing. Why? Because markets always favor efficiency.

Lower systemic friction will also expand transaction volumes. When price discovery functions improve and transaction costs decrease, more economic activity occurs. This is a bullish trend.

Agentic entities representing consumer actions may squeeze platform profits built on "habit." However, they can simultaneously increase total demand by reducing search costs and improving efficiencies.

Productivity is the Core Variable

The ultimate determinant of optimistic outcomes is productivity. If AI can deliver sustained productivity gains in healthcare, government management, logistics, manufacturing, and energy optimization, the result is abundance for all humanity and a lowering of entry barriers.

Even a sustained 1–2% productivity growth will lead to a significant compounding effect over a decade.

The macroeconomic transformation driven by AI has already sparked some of the best investment opportunities in history. This is an area we have spent countless hours researching and continually staying ahead in.

As shown in the graph below, productivity has already started to surge due to AI. In the third quarter of 2025, US labor productivity accelerated, marking the strongest growth in two years:

The pessimistic view assumes that the productivity gains will solely benefit AI model builders and will not translate into broader benefits. The optimistic view, on the other hand, believes that price compression and the formation of new markets will more widely distribute the gains.

Abundance Reduces Conflict, Not Just Costs

One of the least discussed impacts of AI-driven 'abundance' is geopolitics. For most of modern history, wars have been fought over scarce resources: energy, food, trade routes, industrial capacity, labor, and technology. When resources are limited and growth feels like a zero-sum game, competition arises between nations. But abundance changes everything.

If AI substantially lowers the production costs of energy, manufacturing design, logistics, and services, the global economic pie expands. When productivity rises and marginal costs fall, economic growth becomes less reliant on advantaging oneself at the expense of others. This would end wars and could lead to the most peaceful era in human history.

Economic wars are no exception, such as the prolonged trade war we are currently in.

Tariffs are tools from a resource-scarce world to shield domestic industries from cost competition. But if AI causes production costs to collapse everywhere, why would we need tariffs? In a high-abundance environment, protectionism economically becomes inefficient.

History has shown that technological periods of acceleration often reduce global conflict in the long run. The post-WWII industrial expansion decreased the incentives for direct confrontation between major powers.

AI-driven abundance could further accelerate this trend. If energy is managed more efficiently, supply chains are more resilient, and production is more localized through automation, nations become less vulnerable. When economic security rises, geopolitical aggression becomes less rational.

The most optimistic AI outcome is not just higher productivity or a higher stock market index but a world where economic growth is no longer a zero-sum game.

Conclusion: What If the World Doesn't End?

AI amplifies outcomes. If institutions can't adapt, it can amplify fragility; if productivity outpaces the pace of destruction, it can also amplify prosperity.

Anthropic's "unbundling" is a signal that workflows are being repriced, and cognitive work is becoming cheap, marking a clear transition.

But transition does not equate to collapse, just as every major technological revolution has seemed destabilizing in its early days.

The most underestimated possibility today is not utopia, but abundance. AI might compress rents, reduce friction, and reshape the labor market, but it could also usher in the greatest real productivity expansion in modern history.

The difference between a "global intelligence crisis" and "global intelligence prosperity" lies not in capability but in adaptation.

And yet, this world always finds ways to adapt.

Finally, those who can stay objective and follow the process in the current turbulent times are witnessing the best trading environment in history.

Deep Dive Preview: As AI tools like Anthropic showcase remarkable code and workflow automation capabilities, the market plunges into an "AI Doomsday" panic, erasing trillions of dollars in market value. However, this article presents an incredibly insightful reverse perspective: the short-term impact triggered by AI is not a sign of economic collapse but a significant drop in "cognitive costs." By contrasting it with the PC revolution of the 1980s and historical productivity data, the author points out that true "abundant GDP" eras only begin when technology makes knowledge acquisition cheap and abundant. This is not just a labor reconfiguration but a crucial path to geopolitical easing and a global productivity surge.

Original Article Link