The full archive.
Everything from The Ampersand, in order, or sorted by what’s been read most. Search across every essay below.
Searches titles, summaries, and full essay text.
May 2026
What the Machine Cannot Do
After two years of lists about what AI can do, the more valuable list is the other one, and it is the foundation everything durable gets built on.
Where Your Data Goes
The most important AI policy in your company is the answer to one question: what happens to what we type in?
Open Models, Closed Models
The choice between renting intelligence and owning it is older than the technology, and the tradeoffs are the usual ones.
Teaching the Machine Your Business
The model arrives knowing the world and ignorant of you, and closing that gap is most of the real work of implementation.
What an Agent Actually Is
An agent is a model with hands, and hands change the arithmetic of both value and risk.
April 2026
How to Talk to the Machine
The quality of the output is mostly a property of the request, and writing a good request is a learnable skill.
Why the Machine Makes Things Up
Hallucination is a design property, and the businesses that do well treat it as a known cost rather than a recurring surprise.
Sorting the Vocabulary
Five terms cover most of what any vendor will ever say to you, and knowing them is most of the defense.
The Buildings Behind the Intelligence
The intelligence feels weightless, and the infrastructure behind it is some of the heaviest industry on earth.
March 2026
Why Everything Happened at Once
Three slow curves crossed in the same decade, and the crossing explains both the breakthrough and the noise.
The Currency of the Machine
Tokens are how the machine reads, how the vendors charge, and why the meter is worth understanding before the bill arrives.
How the Machine Learns
Training a model is closer to an apprenticeship than to programming, and the difference explains both its power and its blind spots.
The Prediction Engine
A large language model is a machine for guessing the next word, and that turns out to be enough to change how work gets done.
Open Models, Closed Models
The choice between renting intelligence and owning it is older than the technology, and the tradeoffs are the usual ones.
Why the Machine Makes Things Up
Hallucination is a design property, and the businesses that do well treat it as a known cost rather than a recurring surprise.
The Currency of the Machine
Tokens are how the machine reads, how the vendors charge, and why the meter is worth understanding before the bill arrives.
How the Machine Learns
Training a model is closer to an apprenticeship than to programming, and the difference explains both its power and its blind spots.
Seventy Years of Overnight Success
The technology that seems to have arrived overnight has been failing, recovering, and quietly compounding since Eisenhower was president.
What the Machine Cannot Do
After two years of lists about what AI can do, the more valuable list is the other one, and it is the foundation everything durable gets built on.
Where Your Data Goes
The most important AI policy in your company is the answer to one question: what happens to what we type in?
Teaching the Machine Your Business
The model arrives knowing the world and ignorant of you, and closing that gap is most of the real work of implementation.
What an Agent Actually Is
An agent is a model with hands, and hands change the arithmetic of both value and risk.
How to Talk to the Machine
The quality of the output is mostly a property of the request, and writing a good request is a learnable skill.
Sorting the Vocabulary
Five terms cover most of what any vendor will ever say to you, and knowing them is most of the defense.
The Buildings Behind the Intelligence
The intelligence feels weightless, and the infrastructure behind it is some of the heaviest industry on earth.
Why Everything Happened at Once
Three slow curves crossed in the same decade, and the crossing explains both the breakthrough and the noise.
The Prediction Engine
A large language model is a machine for guessing the next word, and that turns out to be enough to change how work gets done.
How to Smell the Hype
A field guide to evaluating AI claims, for owners who will hear a thousand of them this year.
Read before you ever pick up the phone.
Free essays, a few times a month. One field. No sales pitches.
Delivered by beehiiv. No spam, unsubscribe anytime.
- Short essays you can read in one sitting
- How we actually think about AI on Main Street
- No pitches, no funnels. Leave whenever it stops paying