The mind is an agent prototype
Observe how you retrieve memories, decompose tasks, handle interruptions, and form insights. Those cognitive moves are direct blueprints for agent architecture.
Five published essays · newest first
Articles developed
at full length.
The finished writing comes first: developed arguments about AI, philosophy, organizational design, health, and the systems connecting them.
5 published articles · every one ready to readLatest · Mar 31, 2026
From Hierarchy to Intelligence: 2000 Years of Organizational Design, Ending Now
Jack Dorsey and Roelof Botha argue that hierarchy is a 2000-year-old bandwidth constraint. AI world models can replace it. A visual analysis of the most important org design essay of 2026.AIOrganization DesignManagement
Read the essay ↗ Feb 15, 2026
Meta-Drama: The Pulsating Dance of Opposites
Centralization vs. decentralization as the universal plot of reality — from software architecture to physics, mythology to neuroscience.
Philosophy · Systems Thinking · AI Feb 10, 2026
Epistemology Beyond Science
What if science is just one strategy of knowing? A review of Experimental Epistemology and why ML operates above science, not within it.
Epistemology · AI Feb 1, 2026
Token Monism: From Dualism to Unified Abstractions
Why code/data and file/folder dualisms are false — and how a unified token architecture could change programming.
Philosophy · Programming · AI Jan 20, 2026
The Dopamine–Iron Link Nobody Explains
Why your ADHD, restless legs, and bad sleep might share a single biochemical root cause — with 16 peer-reviewed sources.
Health · NeuroscienceDaniel’s idea atlas · 15 theses
Three fields.
One evolving map.
Finished essays sit beside working theories, but never pretend to be the same thing. The status on every card shows whether the thought is published, being tested, or used as a design principle.
AI systems
How intelligence acts.
What changes when intelligence becomes infrastructure?
Open-ended science needs recursive agents
Scalar loops automate problems with a score. Problems without one metric need recursive agents, cross-domain synthesis, and human judgment.
World models can replace coordination layers
When intelligence makes execution cheap, permanent information-routing hierarchies give way to shared world models, capability pools, and temporary problem owners.
A company brain is more than search
An organizational brain should remember entities, decisions, relationships, and changing context—not merely retrieve passages from a document index.
Causal graphs are the missing memory layer
Retrieval finds what was written. A causal graph preserves why a decision happened, what depends on it, and what changed, so an agent can continue instead of restart.
Philosophy
How reality becomes legible.
What kind of knowing produces a useful world?
Reality is a pulsating dance of opposites
Centralization and decentralization are not isolated metaphors but recurring structural forces across software, politics, mythology, mind, and life.
Code and data are one substance
Code/data and file/folder are convenient human dualisms. To an AI, they are mutable token structures whose identity should come from relationships, not location.
LLMs can become a second organ of knowing
First-principles derivation is structurally bounded. Retrieval, synthesis, critique, provenance, experiment, and intervention can turn LLMs into a complementary epistemic layer.
Philosophy is an operating system
A model of time, progress, nature, and human value is not decoration. It changes what you notice, remember, and choose to do.
Choose a positive picture of the future
The same stream of facts can sustain decline or possibility. Choosing an open, future-positive frame is a practical strategy because it changes which actions become imaginable.
Human systems
How people choose.
How does an idea become a way of living?
People are the hardest—and most interesting—system
Bodies, minds, relationships, and societies are less determined than physical systems. Treating people as the central problem changes what is worth learning and building.
Writing is not reporting thought; it creates thought
Writing before speaking forces intuition into structure. It exposes contradictions, clarifies intent, and makes difficult conversations faster and more honest.
Values come before strategy
A plan can be locally rational and still contradict the life you want. Trace choices back to explicit values before optimizing the path.
Aphantasia can act like debugger mode
When visual shortcuts are unavailable, thinking becomes more sequential and inspectable. That slower propositional mode can reveal the machinery behind memory and reasoning.
Concepts should organize facts, not overwrite them
Good sense-making lives between forcing every fact into one theory and letting facts fragment without relation. The map must remain revisable by the territory.