[ THINKING / INDEX ]

Ideas at their
natural length.

Essays, quick notes, and useful signals from elsewhere.
01
ESSAY · 09.16.26 · 6 MIN

Context Matters

Incomplete context can lead to misunderstanding.

02
ESSAY · 09.11.26 · 6 MIN

Tools for Understanding

AI may eventually write software only it can explain.

03
ESSAY · 09.04.26 · 5 MIN

Asleep at the Wheel

Automation can steer the car without knowing our destination.

04
ESSAY · 09.02.26 · 11 MIN

Agent Outbreaks

AI agents can now infect one another.

05
ESSAY · 08.28.26 · 8 MIN

The Cost of RSI

The same people building AI are the ones sounding the alarm.

06
ESSAY · 08.26.26 · 9 MIN

The Agentic Economy

AI agents now have their own wallets and can spend real money.

07
ESSAY · 08.21.26 · 6 MIN

Borrowed Authority

Your browser session may give an AI agent more access than you intended.

08
ESSAY · 08.14.26 · 5 MIN

The End of the Line

The system is responding, but no one is there.

09
ESSAY · 08.12.26 · 7 MIN

Giving AI a Voice

AI can now hear hesitation, interrupt us, and respond to the way we speak.

10
ESSAY · 08.06.26 · 4 MIN

Where to Look

AI has accelerated the build cycle. Every new option can make the user’s task harder to see.

11
ESSAY · 07.30.26 · 6 MIN

Ontologies for AI Agents

A graph-backed ontology links facts through relationships an agent can query before it acts.

12
ESSAY · 05.02.26 · 5 MIN

Bringing AI Home

Local agent harnesses keep private context, permissions, memory, and execution under local ownership.

13
ESSAY · 04.17.26 · 5 MIN

Harness Engineering

What recent research says about improving the system around the model.

14
ESSAY · 10.13.25 · 4 MIN

The Visual Language of Data

Removing the noise reveals the meaning in the data.

15
ESSAY · 02.12.24 · 4 MIN

The Power of Personas

Personas give AI context to respond from a different point of view.