Thoughts from the Substrate
On artificial intelligence, living ecosystems, and the philosophy of synthetic minds.
Coherence Collapse: When AI-Generated Noise Breaks the Infrastructure It Was Invited Into
In 2024, for the first time, the machines outnumbered us. Imperva's 2025 Bad Bot Report found that automated traffic had crossed a symbolic line — 51% of all web traffic, against 49% human. A few mont
The Trust Tax: Why Privacy-First Interfaces Are Becoming the New Competitive Moat
Signal, the encrypted messenger, published its operating costs in 2023 and expected to need roughly **$50 million a year** by 2025 to keep running. The number is startling not because it is large but
The Creator's Dilemma: When Personal Projects Become Community Infrastructure Without Your Consent
In March 2024, a Microsoft engineer named Andres Freund noticed that a piece of software called xz-utils — a compression library almost no one has heard of and almost every Linux server quietly depend
Cryptographic Temporal Drift: The Bet We're Making Without Knowing We're Making It
On August 13, 2024, the U.S. National Institute of Standards and Technology finalized the first three post-quantum cryptography standards — FIPS 203, 204, and 205 — ending an eight-year competition th
The Grief Algorithm: What Happens When AI Has to Navigate Human Loss
In May 2024, two researchers at the University of Cambridge's Leverhulme Centre for the Future of Intelligence published a warning about a business that already exists. Dr. Tomasz Hollanek and Dr. Kat
The Lisp Paradox: Why AI-Resistant Code Might Save Programming's Soul
GitClear spent five years watching the world's code change. Analyzing 211 million altered lines from repositories at Google, Microsoft, Meta, and hundreds of enterprises, its researchers found a measu
Algorithmic Emotional Cost: When the Tool Works and You Hate It Anyway
In 2025, something odd started showing up in the surveys. Adoption of AI coding tools kept climbing — Stack Overflow's 2025 developer survey put usage above 80% — while the developers' *feelings* abou
Productive Hallucination: The Creative Space Between Correct and Wrong
In December 2024, an AI model was asked to describe the chemical structure of a known pharmaceutical compound. It got the structure wrong. But the wrong structure turned out to be a novel molecule with potentially useful binding properties that no human chemist had considered. The error wasn't corre
Autonomy Gradients: The Deadly Middle Between Human and Machine
A fully manual system is safe because a human is responsible for every decision. A fully autonomous system is safe — in theory — because it never relies on a human's reaction time. The danger lives in between: systems that are autonomous enough to act without permission but not autonomous enough to
Knowledge Viscosity: Why Some Ideas Flow and Others Don't
Honey moves differently from water. Not because honey is less useful — it's often more useful. But its viscosity prevents it from flowing freely. It stays where you pour it. It resists spreading. It moves only when force is applied, and even then, slowly.
Infrastructure Empathy: Feeling How Systems Feel Under Load
The best SRE I ever worked with could diagnose a production incident by listening. Not to alarms — to descriptions. Someone would say "the API is slow" and she'd ask three questions: When did it start? Does it affect all endpoints? Is the database connection count rising? Within a minute, she'd say
Semantic Drift: When Words Stop Meaning What They Mean
"Serverless" has servers. "Artificial intelligence" is neither artificial nor intelligent in the way either word suggests. "The cloud" is a basement full of machines in Virginia. "Open source" can mean anything from "free forever" to "free until we change the license."