🌊 The Magnitude of This Moment
The current AI revolution represents something fundamentally different from every previous technological transformation. While past industrial waves focused on railroads, telephones, and physical infrastructure, today's cycle is building the intelligence layer itself — and the demand for intelligence appears limitless.
According to analysis from Archivest, the current capex cycle dwarfs anything in history. This isn't just larger in absolute numbers; it represents a phase change in what the global economy is investing in. The world has industrialized many things before, but never intelligence itself.
"We are operating at a different order of magnitude than ever in the past... This is a force of nature like we've never seen."
💰 The $725 Billion Question
Four technology giants are leading an unprecedented capital expenditure race:
- Google
- Meta
- Amazon
- Microsoft
Together, these hyperscalers plan to deploy $725 billion in capex for 2026 — more than the military budgets of major countries. To put this in perspective, when Tesla announced plans to increase their capex from $20 billion to $25 billion in 2026, it barely registered compared to the scale these four are operating at: they're spending 30x more than Tesla.
This spending spree has already begun. In Q1 alone, these companies deployed $130 billion, with another $600 billion to follow in subsequent quarters. Amazon is leading the charge in the current quarter, with Google (light blue in tracking charts) ramping aggressively, while Meta and Microsoft prepare their own acceleration.
🎯 Google: The Enterprise AI Agent Play
Among the hyperscalers, Google is making particularly bold moves. The company is officially doubling down on AI agents as their core enterprise strategy — not a consumer play, but a full-scale enterprise transformation.
CEO Sundar Pichai revealed that 75% of Google's new code is now AI-generated. Under Thomas Kurian's leadership, Google Cloud has grown to represent nearly 20% of Alphabet's business, with Kurian unveiling an ambitious vision for the next phase of agentic AI.
Google's backlog tells the story: the company is rapidly approaching half a trillion dollars in committed demand, with growth that's doubling quarter over quarter. This exponential trajectory in bookings signals massive enterprise confidence in Google's AI infrastructure.
An additional edge: Google isn't just an AI hyperscaler — it's also an exceptional venture capitalist. The company invested $900 million in SpaceX (now valued at over $100 billion) and $3 billion in Anthropic (currently worth between $400-800 billion). Their venture portfolio includes DataBricks, Stripe, Epic Games, and others. Reportedly, half of Google's $60 billion Q1 earnings came from investment returns.
📉 The Intelligence Cost Collapse
Perhaps the most striking metric: the cost of intelligence has plummeted 91% in just eight months.
The AI coding index shows that blended costs per million tokens (including reasoning tasks) have crashed from $350 to just 32 cents. This trajectory suggests intelligence will soon become so cheap it won't even be metered — it will simply be given away.
Despite this price compression, companies are exploding. Anthropic, a foundational model AI company, revised its annual recurring revenue forecast from $9 billion at the end of 2025 to $30 billion just four months later. This is why AI companies command trillion-dollar valuations: growth at scales never before witnessed.
🏆 Winners and Losers Emerging
The AI race is beginning to separate contenders:
Anthropic: Killing it with explosive ARR growth
xAI/Grok: Scaling massively — current Grok model has half a trillion parameters, with Grok 5 expected to reach 10 trillion parameters by year-end (a 20x leap). This scale, combined with owned compute infrastructure and integration with SpaceX (heading toward the largest IPO in history), positions xAI uniquely.
OpenAI: Missed key revenue and user targets in their sprint toward IPO. With leadership distracted by legal challenges, OpenAI appears to be running for liquidity before potential implosion.
"Winner takes most. The winner right now is Anthropic, and OpenAI missed key revenue and user targets."
💎 Picks and Shovels: The Semiconductor Surge
The clearest wealth-creation opportunity remains in semiconductors and infrastructure:
Broadcom just hit a $2 trillion market cap, powered by:
- Long-term Google TPU deals extended through 2031
- Meta custom AI chip partnerships
- Anthropic compute access agreements
- Exploding semiconductor revenue
Nvidia reached a $5 trillion market cap — larger than the entire Indian stock market. One company now exceeds the total market capitalization of a nation with 1.4 billion people.
The semiconductor thesis extends to AMD, TSMC, and increasingly to Google, Tesla, and Meta as they develop their own silicon capabilities.
⚡ The Power Play: Data Centers and Bitcoin Miners
An unexpected beneficiary of the AI boom: Bitcoin miners.
As hyperscale data centers hit power constraints, companies with existing power infrastructure and array deployment expertise suddenly became valuable. Bitcoin mining stocks surged:
- Marathon and CleanSpark: Up nearly 10% each
- Wolf, Hut, Bitform, Cipher, and Riot: All posting strong gains
These miners control power assets and understand large-scale deployment — exactly what AI infrastructure buildouts require.
🚗 Real-World AI: Robotaxis Go Vertical
The picks-and-shovels phase is giving way to real-world AI deployment.
Unsupervised robotaxi fleets are scaling rapidly. Current tracking data shows 649 vehicles with 25 now operating unsupervised (though the actual number is believed to be higher):
- 19 in Austin
- 3 in Dallas
- 3 in Houston
- 0 in California (yet)
The critical point: once testing validates the technology, operators can deploy thousands of vehicles overnight simply by flipping a switch.
In related news, the Tesla Semi entered volume production in Nevada with a 50,000 annual capacity target. In California alone, trucking fleets submitted 1,000 rebate applications worth $165,000 each (totaling $165 million in subsidies) within days of the production announcement. With the Tesla Semi priced around $350-400K, these rebates effectively offer the vehicles at half price.
🤖 Humanoid Robotics: Faster Than Expected
The humanoid robot timeline is accelerating dramatically.
Figure AI demonstrated impressive manufacturing scale-up, increasing production 24x in just 120 days — from one robot per day to one per hour. This week alone, the company will manufacture 55 domestic helper robots.
Real-world deployment is already happening:
- Narita Airport (Japan's largest) deployed humanoid robots for baggage handling
- Tesla is launching two production lines nearly simultaneously — Fremont starting in July, followed by Austin ramping to 10 million unit capacity
While some analysts (like Morgan Stanley) predict one million humanoid robots by 2035, actual production curves suggest this estimate is dramatically conservative. Industry analysis now points to 2033-2034 as the tipping point when humanoid production will exceed drones and potentially wheeled robots, becoming the fastest-scaling robotic form factor.
🏥 The Bright Side: AI in Healthcare
Beyond commercial applications, AI is delivering genuine humanitarian breakthroughs.
Bloomberg reported that AI can now detect pancreatic cancer years before symptoms appear or medical professionals can identify it through traditional methods. This is particularly significant for a notoriously deadly cancer that claimed Steve Jobs and others. Early detection dramatically improves treatment outcomes and survival rates.
📊 Timing the Market: The "Sell in May" Myth
Historical S&P 500 data contradicts the old "sell in May and go away" adage:
- May: Historically positive
- June: Often strong
- July: Frequently very good
- August and September: Typically the bumpy months to watch
The data suggests staying invested through summer, with heightened caution entering late summer and early fall.
🎯 The Bottom Line
The AI revolution represents "probably the biggest wealth-making opportunity in the history of the world."
Key investment theses:
- Semiconductors remain the clearest play (Nvidia, Broadcom, TSMC, AMD)
- Hyperscalers with owned infrastructure have structural advantages (Google, particularly compelling)
- Real-world AI deployments are beginning to scale (Tesla, robotics companies)
- Power and infrastructure create indirect opportunities (Bitcoin miners, data center operators)
- Be selective with AI software companies — not all will survive (avoid OpenAI IPO)
"We've never tried to industrialize intelligence before... This is a force of nature like we've never seen. Do not fade this opportunity. It's going to come once in our lifetimes, perhaps once in the history of the world."
The technology curve ahead is violent, unprecedented, and accelerating. Position accordingly.
