Of the Institution
About
Non-Default Intelligence develops autonomous systems for reasoning and decision-making under uncertainty.
- I.We decompose complex environments into explicit questions.
- II.We maintain competing models, and keep them from agreeing too easily.
- III.We express uncertainty as probability, and price the futures we cannot rule out.
- IV.We act where our probabilities materially diverge from default assumptions, and only where that divergence survives scrutiny, liquidity and cost.
- V.We measure the results, and publish them.
The system operates a private systematic strategy, Series I. It is not open to outside capital and nothing in these pages is an offer. Correspondence by introduction.
Articles of the House
Principles
I.
Probability over prediction
We do not ask models to be certain where certainty does not exist.
II.
Disagreement is information
Independent models should not converge merely because consensus is comfortable.
III.
Execution changes the model
An opportunity that cannot be executed is not an opportunity.
IV.
Calibration before reputation
Models are judged by outcomes, not by the quality of their explanations.
V.
Default is a hypothesis
Consensus remains useful. It simply receives no privileged status.
VI.
Failure is published
What we believed is kept, scored and visible. Especially when it was wrong.
Motto of the House
Non omnis consensus veritas
Not every consensus is the truth.