Introduction

Welcome to Moby Market Intelligence’s Investor Letter, our monthly note to you, our invested subscribers, on what we’re seeing in the markets and where we’re heading next. Some months it might be a tech sector running especially hot, or the existentialism of regional banks, or it could even be a big-brained look at the world of fast casual dining (were you thinking crypto?). We’ll walk you through the macro backdrop shaping sentiment, question that sentiment, then drill into the names that actually move the story. Think of it as our portfolio meeting, minus the jargon and ego… just the clear, candid Moby view of how we’re positioning our thinking and why it matters now.

The Generative Leap and the Investor Opportunity

This is not your grandfather’s tech cycle. Grandpa had dot-coms; you have data centers that consume more power than Belgium.

Artificial Intelligence, and especially the rise of Generative AI with its Large Language Models, has triggered a structural shift that goes far beyond another upgrade to the tech stack. What we are witnessing is a full-scale economic reorganization disguised in the cloying costumes of not-so-casual startups. Every few decades, the market gets a new operating system, and this one runs on machine learning, GPUs, and an alarming amount of investor caffeine.

Previous booms came and went. The dot-coms collapsed, NFTs are dead and buried, and the metaverse now lives somewhere between memory and meme. AI, however, looks less like a passing craze and more like a new industrial revolution. The phrase “this time is different” still gives every seasoned analyst a mini-stroke, but in this case, it might actually be accurate.

AI behaves the way energy once did. It infiltrates every sector, reshapes cost structures, and demands new infrastructure. The scale of capital flowing into chips, power grids, cooling systems, and data hardware is staggering. This surge is rewriting what investors mean by “defensive positioning.” For anyone paying attention, the AI wave represents a new productivity frontier.

No one can predict the exact outcome. Maybe we end up with self-trading AI funds, or maybe regulators step in before that happens. What matters for investors is that the winners will understand both the code and the copper that power it. Energy, utilities, and materials are no longer the boring parts of the portfolio. They are the essential inputs of the machine that runs everything else.

Defining AI for the Investor

Understanding the Layers of Value

AI gets talked about like it’s one big idea, but in practice it’s a stack of smaller, weirder ones held together by optimism and electricity. Each layer makes money in its own way… and some of them actually make money on purpose.

At the base are the Large Language Models, those industrial engines of the new economy. They eat data, burn through electricity, and cost so much to train that they make private equity fees look humble. Building one is like constructing a nuclear plant that occasionally writes poetry. Only a handful of firms can afford to do it, and even fewer can do it without accidentally teaching the model to swear. And they also need literal nuclear plants.

Above that sits Agentic AI, the point where the machines stop waiting for instructions and start scheduling meetings. This is the version of AI that can draft an email, summarize an earnings call, and politely remind you that it’s already done the job you hired it for. Businesses love it because it delivers scale without the drama of payroll. Labor economists are still in denial, which is adorable.

The economic potential here is that rare mix of measurable and massive. Analysts peg the global AI market at around $758 billion in 2025, expanding to $3.7 trillion by 2034. That’s a growth rate of about 19% per year, which means the next decade will look less like an adoption curve and more like a triple black diamond vertical of capital expenditure. Some forecasters think these numbers are conservative, which feels like saying Timothée Chalamet is “kinda famous.”

Right now, North America holds the lead thanks to its combination of research universities, venture funding, and a national religion built around GPUs. But Asia-Pacific is catching up fast. When entire governments decide to make AI the new industrial policy, it becomes a geopolitical sport. Expect the balance of power to tilt eastward, where data is abundant, regulation is light, and electricity is cheap enough to turn theory into infrastructure.

For retail investors, the key insight is that the AI story does not stop at the software layer. The real value sits in the systems that keep the algorithms alive: energy, semiconductors, cooling, and everything that prevents data centers from bursting into flames. If software ate the world, AI is currently digesting it. Someone, however, still has to sell the silverware.

Why the AI Story is Important to Retail Investors

AI has become the new market religion. Every fund letter, pitch deck, and CNBC segment features it like a sacred acronym that guarantees alpha. The hype feels familiar because it is familiar. Markets always need a new story to justify expensive stocks, and “AI” currently works better than “metaverse,” “crypto,” “future profits,” or “synergy.”

Retail investors should care for two reasons. First, the numbers are already moving markets. Second, those numbers might be hiding structural risk. The major indices have been dragged upward by a handful of names that sound like the guest list at a semiconductor summit. When ten companies carry the weight of the global bull market, you are essentially putting your chips down on the section of the felt that reads “they never miss earnings and never sneeze.”

Diversification still matters, especially when everyone claims it does not. The concentration in AI-related equities makes the whole system fragile. A short circuit in one stock can trip the circuit breaker across an entire index. The smart play might be staying exposed, but spread it across the broader infrastructure that benefits from the same themes of power, cooling, connectivity, and logistics.

Investors also need to remember that optimism is not a risk-management strategy. If the billions currently flooding into AI infrastructure and model development fail to translate into real revenue, valuations will notice, eventually. A correction would sting, but it would also be healthy. Markets with realistic pricing are easier to enter, and actual dips are cheaper entry points than euphoric peaks.

Private equity has also rediscovered retail investors, which is rarely a heart-warming development. When big pools of capital start “democratizing access,” it usually means institutional money is drying up. Offering “exclusive” AI exposure to the public can look a lot like selling umbrellas after it starts raining. What might feel like cautious cynicism is actually self-defense.

Exuberance always feels good until liquidity evaporates. History shows that broad access and high valuations tend to arrive together, and that combination has a habit of ending with finger-pointing and an alphabet soup of post-mortem investigations. Balance the excitement with skepticism, keep a shopping list for the pullback, and remember that bubbles only look obvious after they pop.

The short version: AI is too big to ignore, too hot to chase recklessly, and too early to short. The job of the retail investor is to enjoy the upside without volunteering as exit liquidity for the institutions that got there first.

AI Touches Everything 

Disruption Mapping Across the Economy

At Moby, we look at AI the same way economists look at oxygen. It’s everywhere, it’s invisible, and most of the market is still underestimating how much of it they’re inhaling. Every business that runs on electricity is already an AI story; the only question is whether investors have noticed yet.

We break it into two categories: “The Obvious Bets” and “The Hidden Transformations.”

The Obvious Bets: Core Infrastructure and Enablers

The first group is easy to find because they won’t stop talking about it. These are the companies selling the picks, shovels, and GPUs powering the gold rush. The global computing stack is migrating from Central Processing Units to Graphics Processing Units, a shift sometimes called Software 2.0. In practical terms, we’ve moved from code written by humans to code learned by machines. That transition has created the first truly physical bottleneck in the digital age: compute capacity.

NVIDIA (NVDA) — ever heard of it? — still dominates both the training and inference markets, minting cash like a tech-sector central bank. Advanced Micro Devices (AMD) remains the most credible alternative for buyers who prefer optionality and lower valuation risk. Broadcom (AVGO) is carving out the high-margin niche of custom AI chips for hyperscalers who want proprietary performance. And TSMC (TSM) quietly manufactures silicon for nearly everyone, acting as the geopolitical bottleneck that keeps both markets and diplomats awake at night.

These companies don’t just sell hardware; they sell speed. In the AI economy, faster training equals faster revenue, and compute time has become the new yield curve. Capital now flows toward processing power the way it once chased interest rates. This cycle is still early, and the constraints are very real: power, cooling, materials, and supply-chain precision. The players solving those problems are building the next generation of market infrastructure, and that infrastructure is far from fully priced in. We’ll be talking about those names a lot in the coming weeks and months.

The Non-Obvious Transformation: Hidden AI Exposure

The real fun starts in the second group. These are the companies that never show up on AI stock lists but quietly use it better than anyone. They turn hype into operating leverage while the market keeps rewarding press releases instead of execution.

Walmart (WMT) is the perfect example. On paper, it is a retailer. In practice, it is an algorithm that sells detergent. Predictive models choreograph its supply chains, forecast demand, and personalize promotions, all while letting customers check out through ChatGPT. It does not trade like an AI stock, but it runs like one, and that difference matters.

The same shift is spreading everywhere. Freight carriers reroute shipments before storms hit. Hospitals anticipate diagnoses before symptoms show. Manufacturers fix machines before they break. The payoff is already visible in earnings: cleaner margins, faster turns, and CFOs suddenly discovering they have been “AI-focused” all along.

These are the sleepers hiding in plain sight. They are capturing the efficiency premium without paying the valuation penalty. While the crowd chases the next shiny model, these operators are quietly building moats with code.

AI is now infrastructure. The winners will be the ones who treat it like plumbing, not marketing. The smart investors are already tracing the wiring, looking for balance sheets where the lights are quietly on.

Risks and Opportunities

Markets are stuck in a holding pattern. Tariffs, politics, and general investor fatigue are keeping risk appetite low. Too much money sits in too few AI names, which is fine until gravity kicks in. If those leaders stumble, the entire index goes with them. Diversification into assets that don’t all move with Big Tech is the practical move. There’s opportunity here, but it rewards patience over panic.

The Digital Transformation

A Fundamental Business Model Shift

Digital transformation means exactly what it sounds like: replacing human drag with digital efficiency. AI is the operating system of that shift. It runs workflows faster, trims costs, and reduces decision time from days to seconds. But this is not an optional upgrade. Staying competitive now requires AI fluency. Everyone else becomes a case study.

1.

What Does it Mean for Operators?

AI adoption forces companies to rebuild from the inside out. The “Agentic Organization” model replaces siloed structures with compact, outcome-driven teams managing swarms of AI agents. A few supervisors can now oversee an operation that once needed hundreds of people. The math is obvious: faster execution, fewer layers, no burnout.

2.

What Does it Mean for Consumers?

For customers, this shift shows up as convenience that feels borderline psychic. AI processes your habits in real time and serves up exactly what you were about to look for. The end result is stickier users and higher margins. Netflix and Amazon have arguably been pretty successful with it.

3.

What Does it Mean for Investors?

The edge now belongs to the companies that can actually make AI pay for itself. The marketing phase is over. The winners are the ones producing measurable returns from real deployments, not from PowerPoint decks. Most firms are still pretending. The small number doing it for real are where the opportunity lives.

Navigating the AI Hype Cycle

Yes, the hype is real. No, that doesn’t make it wrong. Every tech cycle starts with too much optimism and ends with a handful of survivors who make all the money. We’re somewhere in the middle at this point. There’s plenty of noise, but also plenty of signal if you can separate them.

Conclusions

The Energy-AI Nexus

AI doesn’t run on magic, or even chips. It runs on power, and we don’t have enough of it. Data centers are devouring electricity like frat boys at an open bar. BloombergNEF projects usage will top 1,200 terawatt-hours by 2035, and that’s before every enterprise server farm gets an AI upgrade.

The punchline is that utilities and infrastructure companies are suddenly growth stories. Not the flashy kind, but the “quietly compounding cash flow” kind. Big Tech has already figured it out. Google, Amazon, and Meta are locking in long-term energy deals like they’re building their own power grids… because they basically are. The chip shortage was fun and all, but the real constraint is juice.

Investment Opportunities in AI Energy Enablement

Follow the amperage, not the app store. Utilities near data center corridors are the stealth winners, turning stable demand into steady growth. Cooling tech, energy storage, and grid optimization are where the smart capital is migrating. Companies that keep data centers cool, stable, and online are the new infrastructure elite.

Grid stability isn’t sexy, but it pays. Battery storage firms, demand-flexibility operators, and cooling-as-a-service providers are becoming critical to AI’s backbone. Meanwhile, everyone still chasing “AI software multiples” is playing musical chairs with overvalued tickers. The real money is in keeping the lights on for the companies running the models.

AI is the headline. Energy is the business model. The future of AI looks less like Silicon Valley and more like a construction site with a substation attached.

Energy as a Springboard for Future Moby Research Content

Moby’s next wave of research is going beneath the hype, to the physical infrastructure holding the entire AI economy together. We’re talking about the cooling systems that keep GPUs from melting, the nuclear plants that quietly hum in the background, and the grid technology that stops everything from flickering when demand spikes.

Copper is the unsung hero of this whole buildout (and all buildouts if we want to get technical). It’s in the wires, turbines, batteries, and cooling systems that keep digital infrastructure alive. Supply is tight, demand is rising, and no amount of AI optimization can mine more copper. It’s the choke point in the next great trade.

To track the companies building the foundation of this new energy ecosystem, Moby will launch the Moby Energy Infrastructure Index (MEII). It’s a curated portfolio of firms generating, transmitting, and cooling the AI economy. We’re talking real assets, real earnings, none of the bubble-era valuation gymnastics. The goal is simple: own the picks and shovels of the AI gold rush without buying into the mania.