Cumulative thesis
Reasoning Made Legible
What does a model know, and how can we verify it?
Language models reason probabilistically. Computational systems execute explicit structure. The useful frontier is learning where each form of intelligence should carry the load.
Where we land
We do not need machines to think exactly like us. We need their reasoning to become observable enough to inspect, challenge, and trust where the stakes demand it.
Essays in this thread
4 essays
What the Model Thinks But Doesn't Say
Anthropic found a global workspace inside Claude: a silent, evolving set of unspoken words the model thinks with. Learning to read it may matter more for the agent economy than the next benchmark record.
Essays
The Idiot Index of Intelligence
Strip any product down to its atoms and the gap between that floor and its price is pure waste. What happens when you point that lens at a thought?
Essays
Ontology, Context and Semantic Data Layers
Ontology defines what a business means. The context layer decides how agents retrieve, permission, verify, and use that meaning.
Essays
The Future of Human-AI Collaboration: Beyond Replacement to Partnership
Exploring how the next phase of AI development focuses on augmenting human capabilities rather than replacing them entirely.
Essays