Recent Work

The Mycelium of Meaning

How Language Models Grow, Fail, and Reshape the Question of Understanding

Coming Soon on Amazon

Artificial intelligence did not arrive fully formed. It grew. For more than seventy years, researchers have tried to teach machines how language works — and each attempt solved a real problem while exposing another boundary. The Mycelium of Meaning traces that history and asks a harder question than whether machines are becoming intelligent: what do their successes and failures reveal about language, comprehension, and us?

This is a book about the development of machine language systems — from rule-based and statistical methods through neural networks and transformers to retrieval, agents, and the systems that surround them. It reads that history as a way of thinking clearly about a set of distinctions that keep getting blurred.

The Central Inquiry

The book's subject is not "AI" in the generic sense. It is a sustained inquiry into the difference between what these systems do and what we are tempted to believe they do — the distinctions between:

syntax and meaning
probability and understanding
representation and truth
machine capability and human intent
intelligence and relationship
simulation and lived meaning

Why Mycelium

The governing metaphor is the mycelium — the underground network of fungal threads that connects a forest. It is the book's way of holding a claim about meaning itself: that meaning is relational, distributed, contextual, and alive — generated through connection rather than contained in isolated symbols.

A rule-based system captured syntax but not meaning. A statistical system predicted linguistic habits without establishing understanding. A neural network mapped language into geometry without turning coherence into truth. Each era achieved something genuine, and each mistook a necessary part for the sufficient whole. The mycelium is what remains: not a single symbol that carries meaning, but the living web of context, use, and consequence in which meaning grows.

Meaning emerges through relationships of context, use, interpretation, and consequence — while responsibility remains with those who design, deploy, evaluate, and live with these systems.

Symbiosis, Not Replacement

The Mycelium of Meaning ultimately argues for a human–machine relationship built on symbiosis rather than replacement. Distinct intelligences can produce meaning together without becoming the same thing. But symbiosis is not surrender: human intention, judgment, responsibility, and meaning-making remain central.

This is neither a prophecy of machine consciousness nor a dismissal of what these systems can do. It is a disciplined account of capability without mythology, collaboration without sameness, and responsibility without the illusion of complete control.

Read a Sample

Chapter 1, The Mechanical Dream, follows the rule-based era — from the 1954 Georgetown–IBM translation demonstration to ELIZA and SHRDLU — and arrives at the axiom the rest of the book stands on: meaning cannot be reduced to syntax.

Read Chapter 1 → Essays