Tech Is Laying Off Workers and Crowning AI Winners. Here's Where the Money Is Actually Going.

Prediction markets are painting a remarkably clear picture of what's happening inside the tech industry right now, and it looks a lot like the early stages of a Gold Rush. Not the kind where everybody gets rich, but the kind where a few miners strike it big while thousands go home broke. The real money? It went to the people selling pickaxes and shovels.

Let's start with the numbers that tell the story.

Betting markets currently price an 83% chance that tech layoffs in 2026 will exceed 2025 levels. At the same time, the race to build the best AI model is tightening dramatically. Anthropic (maker of Claude) sits at a 57% probability of leading AI by December 2026. Google trails at 24%. OpenAI, which just two years ago seemed untouchable, has fallen to just 11%. Elon Musk's xAI rounds out the field at roughly 10%.

Meanwhile, the SpaceX IPO carries a 75% probability of happening by July 2026, potentially injecting a burst of excitement into tech investing. And there's a 39.5% chance that U.S. unemployment tops 5% by 2027, a level that would signal real economic pain beyond just the tech sector.

Put these signals together and a pattern emerges: the tech industry is going through a massive reshuffling. Capital is concentrating into fewer AI winners while the broader tech workforce shrinks. Economists have a term for this: creative destruction. Old business models break down so new ones can take their place. The trouble is that the "destruction" part tends to arrive before the "creative" part finishes building.

The Self-Reinforcing Cycle

This reshuffling feeds on itself in a way that's worth understanding, because it explains why the trend is likely to accelerate rather than reverse:

AI models get more capable, allowing companies to automate tasks previously done by mid-level software engineers, customer support teams, and data analysts.
Companies lay off workers in those roles, which is already showing up in the 83% layoff probability.
The savings from those layoffs get redirected into AI infrastructure spending, because falling behind in the AI race means losing to competitors who didn't.
That increased AI spending makes models even more capable, which starts the cycle over again.
The winners consolidate. The gap between the leading AI labs and everyone else widens, which is exactly what the prediction market odds reflect: Anthropic pulling away at 57% while OpenAI sinks to 11%.

This is Ray Dalio's "paradigm shift" framework in action. The old model of tech employment, where companies hired armies of engineers and paid them handsomely to build and maintain software products, is breaking down. The new model concentrates value in a small number of AI platforms and the physical infrastructure that powers them.

Shovels, Not Gold

During the California Gold Rush of 1849, the merchants who sold picks, shovels, and denim pants made more reliably than most miners. The same logic applies to AI. We don't know for certain whether Anthropic, Google, or some dark horse will be the dominant AI platform in 2028. But we know with near-certainty that all of them need chips, networking equipment, cooling systems, and electrical power.

That's the infrastructure thesis, and it's where the strongest investment signals are pointing.

**VRT (Vertiv)** gets the highest conviction call as a Strong Buy at 82% confidence. Vertiv makes power management, thermal management, and cooling systems for data centers. Think of them as the company that keeps the AI factories from overheating, literally. As AI models get larger and more compute-intensive, the amount of heat generated per server rack is skyrocketing. Cooling has become the critical bottleneck. Whether Anthropic, Google, or OpenAI wins the AI race, every single one of them needs Vertiv's equipment. The company scores 88 out of 100 on infrastructure relevance, with data center equipment comprising essentially their entire business.

**ASML** is a Buy at 78% confidence and carries the highest infrastructure relevance score of 92. ASML holds a literal monopoly. They are the only company on Earth that makes the extreme ultraviolet (EUV) lithography machines needed to manufacture advanced AI chips. Every cutting-edge chip from NVIDIA, AMD, Broadcom, or any custom design has to be made on equipment that only ASML sells. It's like owning the only bridge across a river that everyone needs to cross.

**PWR (Quanta Services)** is a Buy at 76% confidence. Quanta builds the electrical transmission lines, substations, and grid connections that data centers require. AI data centers are enormous power consumers, and increasingly the bottleneck isn't getting enough chips but getting enough electricity to the building. An estimated 30-40% or more of Quanta's new orders are tied to data center and renewables electrical work.

**ETN (Eaton)** is a Buy at 74% confidence. Eaton manufactures switchgear, uninterruptible power supply systems, and power distribution units that every data center needs. They compete with and complement Vertiv in the power management space. Their diversification across aerospace, vehicles, and industrial segments provides a cushion if AI spending ever slows, though it also means less pure-play exposure to the AI buildout.

**ANET (Arista Networks)** is a Buy at 73% confidence. Arista makes the high-speed networking switches that link GPU clusters together for AI training and inference. Think of their products as the nervous system of an AI data center. As AI models scale and require more distributed computing power, the network bandwidth between chips becomes a critical performance bottleneck. Arista's 400G and 800G switches are the connective tissue that makes large-scale AI possible.

**EQIX (Equinix)** rounds out the infrastructure plays as a Weak Buy at 65% confidence. Equinix operates carrier-neutral data centers and interconnection hubs globally. They benefit from the broad trend of companies needing AI-adjacent infrastructure without building their own. The lower confidence reflects two concerns: hyperscalers like Amazon and Google are increasingly building their own data centers rather than leasing, and as a real estate investment trust (REIT), Equinix is sensitive to interest rate movements.

The Platform Players

Beyond pure infrastructure, a few of the AI platform companies themselves look attractive.

**AVGO (Broadcom)** is a Buy at 75% confidence. Broadcom designs custom AI accelerators, including Google's TPU chips, along with the networking semiconductors that connect data centers. They sit at the intersection of infrastructure and platform, profiting from AI buildout regardless of which model wins. Their acquisition of VMware also positions them to benefit as companies rationalize their cloud spending during the restructuring.

**AMZN (Amazon)** is a Buy at 72% confidence, primarily because of AWS, Amazon's cloud computing division. AWS is a picks-and-shovels play hiding inside a retail company. It hosts AI workloads for Anthropic (Amazon has invested over $4 billion in the company), startups, and enterprises. The creative destruction pattern means more companies will rent AI computing power rather than build it themselves. Amazon has also already been through its own layoff cycle, which means they're ahead of the restructuring curve and positioned for better margins.

**GOOGL (Google)** gets a Weak Buy at 62% confidence. Google's 24% probability of leading AI by December 2026 could represent a value opportunity if the market is underpricing their Gemini AI model. Google has unmatched data assets, proprietary TPU chip infrastructure, and distribution through Search, Android, and YouTube. But the prediction markets are clearly telling us Anthropic is the favorite, and Google's traditional advertising business could face disruption from AI-native search alternatives. This is more of a speculative position and a hedge against the possibility that the "Anthropic wins everything" narrative turns out to be wrong.

Why This Matters for Your Money

You don't need to trade individual stocks for this pattern to affect you. If you have a 401(k) or index fund, you already own most of these companies. The question is whether your portfolio is weighted toward the old tech employment model (broad-based SaaS companies, mid-tier software firms) or the new one (AI infrastructure, dominant platforms).

The 83% layoff probability has implications beyond stock prices. If you work in tech or know someone who does, the job market is likely to get more competitive, not less. The roles that survive and thrive will be the ones closest to AI development and infrastructure, the digital equivalent of selling shovels.

And that 39.5% chance of unemployment crossing 5% means there's a meaningful risk this isn't just a tech story. If AI-driven layoffs spread to other industries, it could affect consumer spending, housing, and the broader economy in ways that touch everyone's grocery bills and savings accounts.

The Risks You Can't Ignore

Every signal here comes with real downside risks that deserve honest attention.

For the infrastructure plays, the biggest risk is that many of these stocks have already run significantly. Vertiv, ASML, Arista, and others are trading at premium valuations that assume continued explosive growth in AI spending. If the AI investment cycle peaks or even pauses, these stocks could give back gains quickly.

ASML faces geopolitical risk because its largest customer is TSMC in Taiwan. China export restrictions are already shrinking its addressable market. The semiconductor industry is cyclical by nature, and the current AI spending boom could eventually create overcapacity.

Arista has customer concentration risk, with Meta, Microsoft, and other hyperscalers making up a huge portion of revenue. If any major customer decides to bring networking in-house the way they've brought chip design in-house, Arista would feel the pain.

For Broadcom, the VMware integration could distract management at a critical moment, and the custom chip business depends entirely on hyperscalers continuing to spend at current rates.

Amazon faces antitrust scrutiny, and its retail segment is vulnerable to consumer weakness if unemployment does rise above 5%. Their $4 billion bet on Anthropic could become a liability if Claude loses ground to competitors.

Google faces the most complex risk profile. The DOJ antitrust case could force structural changes. Gemini development has been inconsistent. And there's real brain drain happening as top AI researchers leave for startups.

The SpaceX IPO at 75% probability could provide a temporary sentiment boost for tech investing, but it doesn't change the underlying trend: capital is concentrating, not spreading.

Perhaps the broadest risk is that the 83% tech layoff signal turns out to reflect genuine demand destruction rather than productive restructuring. If companies are laying off workers because business is bad, not because AI is making them more efficient, then the entire thesis flips from "creative destruction" to plain old destruction. The infrastructure buildout only makes sense if someone is willing to keep paying for all that computing power.

Analysis based on prediction market data as of April 15, 2026. This is not investment advice.