Essays
Essays and notes on AI, systems, and the meeting point between technology and Theosophy, exploring memory, identity, consciousness, and the human experience.

July 21, 2026 · 4 min read
The Last Interface: software in the age of AI agents
For fifty years, software has assumed a human on the other side of the screen. As AI agents become the primary operators of software, the interface may stop being the center: capabilities matter more than screens, APIs become the product, and the so-called SaaS apocalypse looks less like a disappearance than an inversion.
Read

July 18, 2026 · 6 min read
Five Cats and the Tenant's Question: the body is the hardware, instinct is the software, so who is the user?
Five cats share one body plan and one bundle of instinct, yet nobody could confuse them. On what nature hands over ready-made, what genetically identical mice and a cloned cat reveal about where individuality comes from, where the sober science ends and my bet on the monad begins, and the same question turned back on myself.
Read

July 17, 2026 · 7 min read
The Store That Forgets Me: my supermarket holds a perfect record of my life and greets me like a stranger anyway
Every receipt is a faithful, consented record of a life, and most grocers throw it away. On the corner grocer who remembered his regulars, the chain that proved memory pays back in 1994, the fear that became an industry's alibi, and why continuity is the courtesy retail stopped extending.
Read

July 16, 2026 · 5 min read
When AI Remembers You Wrong: an AI can get one answer wrong, but a persistent memory can keep reproducing the error
When an AI turns a temporary observation into a lasting belief, the mistake stops being a one-off answer and becomes infrastructure — quietly shaping every interaction that follows. The first of five pieces on continuity without distortion, and an introduction to biographical hallucination.
Read

July 15, 2026 · 14 min read
The Last Editor: after decades of teaching people how to operate software, we are building software that learns how to listen
Every product I worked on had an editor in the middle, asking people to translate intention into fields, screens, and conditions. Language agents reverse that direction of translation — but the editor doesn't disappear, it becomes invisible. On what agentic product design owes users when the system carries the structure.
Read

July 14, 2026 · 9 min read
Diagnosis Was the Easy Part: as machines learn to out-diagnose doctors, the frontier of personal health turns out to be a faithful memory
As language models begin to out-diagnose physicians, the binding constraint in personal health stops being medical knowledge and becomes something older and humbler: a continuous, faithful record of a single life — and fidelity, not storage, is the real frontier.
Read

June 26, 2026 · 15 min read
How I Would Try to Break the Map: Five experiments against a six-paper research programme
The adversarial companion to the monad essay: the cheapest sequence of experiments that could make the six-paper programme smaller, and the concessions written down before the data arrive — where perspective, self-models, relationship, inheritance, and developmental order each fail.
Read

June 25, 2026 · 16 min read
We Scaled the Ape and Expected the Human: Theosophy's developmental cosmology as a heuristic engine for machine learning
A speculative essay on borrowing the journey of the monad as a disclosed source of machine-learning variables: faculty order, perspective supervision, external accountability versus relational reciprocity, lifecycle consolidation, and the mistake of expecting a self to appear from scale alone.
Read

May 26, 2026 · 4 min read
A Cell Is a Neural Network
Reading biology as deep learning: immune systems as fine-tuning, allergies as overfitting, autoimmune disease as alignment failure, DNA as compressed weights.
Read