Systems & Access
The infrastructure, economics, and standards that shape AI access.
AI and Education: Tutoring, Testing, and Who Gets Left Out
In 1984, the education researcher Benjamin Bloom published a finding so striking it became known simply as "the 2 sigma problem." Students who received one-on-one tutoring performed two standard deviations better than students in conventional classrooms. That's the difference between an average student and one in the 98th percentile. The problem was cost: you can't give every student a private tutor. Bloom's paper was essentially a challenge — find a way to deliver tutoring-level instruction at classroom scale.
AI Chip Export Controls: Technology as Geopolitics
In October 2022, the United States did something unusual. It deliberately hobbled its own companies' most lucrative export market. The Bureau of Industry and Security issued sweeping export controls blocking China's access to advanced AI chips, chip manufacturing equipment, and even the expertise of American citizens working in Chinese semiconductor facilities.
Compute Is the New Oil
Training frontier AI models requires computing power that only a handful of organizations can afford. This concentration has consequences that extend far beyond who can build the biggest model.
Cryptographic Post-Economies: Can Token Systems Redistribute AI Value?
As AI automates more of material production, a seductive idea keeps resurfacing: what if cryptographic systems — tokens, smart contracts, decentralized autonomous organizations — could route AI-generated value directly to people, bypassing the corporations that currently capture it?
The Data Center Boom and Who Pays the Energy Bill
Somewhere in rural Virginia, a farmer watches construction crews pour the foundation for a data center the size of several football fields. The county offered a twenty-year tax abatement to attract the project. The power company is building a new substation to feed it. The farmer's electricity bill went up last year to help cover grid upgrades. The data center, when it opens, will employ about fifty people.
Edge AI and the Promise of Local Compute
There is a version of AI that doesn't phone home. No API call, no cloud server, no token meter ticking upward with every query. The model runs on the device in your hand or on the desk in front of you. The data stays where it is. The inference is free — or rather, it's already paid for, bundled into the price of the hardware.
The Environmental Cost of AI
Every time you ask an AI model a question, a server somewhere draws electricity, generates heat, and consumes water to cool itself back down. Multiply that by hundreds of millions of daily queries, add the months-long training runs that created the model in the first place, and factor in the mining, manufacturing, and shipping of the hardware — and you begin to see the outline of an environmental footprint the industry would rather not discuss.
How Technology Gets Distributed: Models That Shape Who Benefits
Every transformative technology faces the same question: who gets to build it, and on what terms does everyone else get access? The answer is never purely technical. It's a design choice — shaped by policy, funding, ideology, and power. AI is no exception. But the choices being made right now will lock in patterns that persist for decades.
Open Source vs. Open Weights
Meta's Llama is often called "open source." It isn't — not in the traditional software sense (Widder et al., 2023). Understanding the distinction matters for who actually controls AI infrastructure.
Pricing Power: How API Costs Shape Who Can Build
API pricing is not neutral infrastructure. It is a mechanism that determines who can build AI applications, what use cases are economically viable, and how value flows between providers and developers. When OpenAI charges $2.50 per million input tokens for GPT-4o and $10 per million output tokens, those numbers encode decisions about who gets access to frontier capabilities — and on what terms.
The Rise of Small Language Models
The prevailing narrative of AI progress runs in one direction: bigger. More parameters, more training data, more compute, more money. GPT-4 reportedly cost over $100 million to train. Google's Gemini Ultra required clusters of thousands of TPUs. The arms race has a clear logic — scale up and performance follows. But a counter-trend is quietly reshaping who gets to use AI, and on what terms.
The API Economy
You don't run GPT-4. You call it. This architectural choice — AI delivered as a service through APIs — shapes who can build what, and who captures the value.
The Frontier Lab Oligopoly
Five organizations control the development of the most powerful AI systems on Earth: OpenAI, Anthropic, Google DeepMind, Meta AI, and Mistral. This is not a diverse marketplace. It is an oligopoly, and the structure of that oligopoly shapes what AI gets built, who benefits from it, and what alternatives can exist.
The Hardware Bottleneck
The AI industry runs on software that can be copied infinitely at zero marginal cost. But it depends on hardware that cannot. The chips that train and run AI models flow through supply chains that are fragile, concentrated, and increasingly contested. Understanding these chokepoints is essential to understanding who controls AI.
The Inference Tax: How API Pricing Creates Perpetual Rent
Every time you call a commercial AI API, you pay rent. Not a fee for a product you now own. Not a license that grants you permanent rights. Rent — paid to the platform owner for temporary access to capability you will never possess.
The Magnificent Seven
Seven companies — Apple, Microsoft, Google (Alphabet), Amazon, Meta, Nvidia, and Tesla — now represent roughly 30% of the S&P 500's total market capitalization. At moments in 2024, that figure exceeded 33%. These seven companies, with a combined market value surpassing $15 trillion, are worth more than the entire stock markets of most countries.
The Redistribution Stack
Some players are actively redistributing AI value outward. Others talk openness while positioning themselves at the center. Here's how to tell the difference.
Who Controls the AI Supply Chain
When you ask a chatbot a question, you interact with a software layer that sits atop one of the most concentrated, geographically fragile, and politically contested supply chains ever constructed. Seven layers of infrastructure connect minerals in the ground to the model on your screen. At each layer, a small number of actors control what gets built, where, and for whom.
Who Owns the Training Data?
Public data trains private models. The internet's collective output — Wikipedia articles, Reddit comments, news archives, academic papers, books, images, code — becomes the raw material for systems worth hundreds of billions.