Over the past decade, I've been developing a framework for auditable, traceable, and composable reasoning—intended to scale across domains like law, science, and strategy. It's not within the LLM paradigm, but complementary: a principles-first approach focused on explainability through generated proof objects called Reasoning DNA Units (RDUs).
The manifesto introduces the theory behind RDUs, the requirement of a a domain-specific language to express them, and open-source prototypes. I’m proposing a communal effort to build this into a shared substrate for trustworthy reasoning systems.
Over the past decade, I've been developing a framework for auditable, traceable, and composable reasoning—intended to scale across domains like law, science, and strategy. It's not within the LLM paradigm, but complementary: a principles-first approach focused on explainability through generated proof objects called Reasoning DNA Units (RDUs).
The manifesto introduces the theory behind RDUs, the requirement of a a domain-specific language to express them, and open-source prototypes. I’m proposing a communal effort to build this into a shared substrate for trustworthy reasoning systems.
Manifesto: https://doi.org/10.5281/zenodo.17180040
Prototype: https://github.com/Eric-Robert-Lawson/OrganismCore