Recent writing

Recent posts

What Will AGI Actually Look Like?
substack / Jul 16, 2026

What Will AGI Actually Look Like?

Whenever people debate AI, they usually think they’re talking about the same thing. They’re not. Most conversations actually mix four completely different paradigms: 0. AI is mostly hype. 1. AI can already perform a larg

Read ->
What Should You Build in the Age of AI?
substack / Jul 15, 2026

What Should You Build in the Age of AI?

Every technological revolution begins with an opportunity and ends with a commodity. At first, the new thing is rare, mysterious, and expensive. A handful of people know how to build it. Investors queue outside. Conferen

Read ->
Archive

Older notes and essays

medium / Jul 10, 2026
BabelTele: When AI Stops Speaking Human

BabelTele suggests that LLMs can exchange far denser messages than human language allows, cutting context costs while creating a new problem: machine conversations we may no longer be able to audit. For the entire histor

substack / Jul 08, 2026
How to Build an Agent That Actually Gets Better Over Time

Most “self-learning agents” are not self-learning. They are agents with a memory file, a dramatic name, and the same tragic ability to make the same mistake next Tuesday. This is not because the idea is bad. The idea is

substack / Jul 06, 2026
The First Herculaneum Scroll Has Finally Been Read

For nearly two thousand years, the Herculaneum scrolls have been playing the most infuriating game in the history of books. They survived. But only by becoming unreadable. When Vesuvius erupted in AD 79, the library at H

medium / Jul 06, 2026
ChatGPT Does Not Pick Sources the Way SEOs Think It Does

A developer inspected ChatGPT’s hidden source-selection data and found that AI search is less about ranking pages and more about recruiting readable, quotable evidence. For twenty years, SEO people have lived inside the

substack / Jun 30, 2026
Mythos and the End of Comfortable Cybersecurity

There are sentences you read once, put the coffee down, and then read again more slowly. This is one of them: “It broke into almost all of our classified systems, not in weeks, but in hours.” That line was attributed to

substack / Jun 29, 2026
Stop Prompting AI. Start Engineering Loops.

There is a new phrase floating around the AI coding world, and it sounds just irritating enough to become important. Loop engineering. Naturally, it has arrived with the usual smell of fresh terminology, LinkedIn smoke,

substack / Jun 28, 2026
Paid Newsletters in 2026: The Paywall Is Not the Product

There was a time when a paid newsletter was a very simple machine. You wrote something. You put half of it behind a paywall. Then you sat there like a man selling umbrellas in the Sahara, wondering why nobody was throwin

substack / Jun 24, 2026
The Great Wrapper Panic

There is a particular kind of business nightmare that arrives very politely. It does not kick the door down. It does not announce itself with a dramatic soundtrack. It simply walks into the room, opens a benchmark, and s

substack / Jun 23, 2026
AI Music Is Not Replacing Musicians. It Is Replacing Spam.

There is a wonderful old fantasy about the music industry. It goes like this: people make songs, listeners decide what they like, critics write reviews, platforms count plays, charts reflect popularity, and everyone invo