A System of Heredity for Intelligence

Institutional learning

Checked inheritance of commitments and capability across generations of neural and non-neural computational systems.

This program studies how a computational descendant can inherit a versioned set of commitments and verified capabilities while using a different executable mechanism. The design separates the inherited specification, the construction plan, the builder, the independent verifier, and the resulting substrate. The proposed generational experiments examine whether useful capabilities and defined constraints survive both succession and changes in computational mechanism. A new artifact receives fresh evaluation and admission rather than inheriting authority merely through its lineage.

Evidence priorities

Make progress measurable.

The initiative is framed around testable questions. Its research direction should be assessed against observed evidence and the limits of each experiment.

• Explicit commitments and capabilities preserved across generations
• Independent evaluation after changes in executable mechanism
• Reversibility and admission of each new descendant

Research goals and proposed mechanisms are not established findings. Product-linked tracks describe the research behind a capability, rather than a claim that its hypotheses have been validated.