Human Source Code

Overview

Human Source Code is a project to build code that describes the human as a multiscale system: from cellular and physiological processes, through cognitive-affective mechanisms, to individual behaviour in a social environment. Each layer carries an explicit state model and a time-step function, and the layers communicate through a shared buffer or event messages, so that modules can be developed independently and the system tested in real time.

Its guiding principle is evidence-by-design: every dependency in the code carries an annotation of its source, confidence level, and scope of applicability. Questions without solid data support remain experimental parameters with default values and sensitivity tests. Knowledge is versioned semantically, and changes carry provenance metadata, so that any simulation run is reproducible with exactly the same models, parameters, and inputs.

The architecture divides into four layers: physiology (homeostasis, energy reserves, rhythms, stress signals); a cognitive-affective layer (perception, memory, learning, emotion regulation); an intentional-decisional layer (goal formation, trade-offs, executive control); and a social and environmental layer (roles, norms, relation networks, context). The couplings between them are explicit: a sleep deficit in physiology degrades working memory; chronic stress modifies the choice rules in the decisional layer. The model does not assume determinism. Uncertain parameters are represented as distributions, and results report confidence intervals and posterior distributions.

The language of the project is engineering, but its aim is not to profile individuals. It is a model of the species: a repository of hypotheses and mechanisms that can be tested, falsified, and swapped. The core stays parsimonious; detail is added only when it improves predictive accuracy or enables an empirical test. A governance layer is built in, with time and step limits, a signed registry of decisions and state changes, and explicit ethical policies, including no profiling of sensitive attributes and consent required for data.

Defined this way, the human source code creates a frame in which knowledge about physiology, psyche, and behaviour is systematically recorded, versioned, and tested, from a minimal working core to richer modules that gradually become encrusted with evidence. A multi-year endeavour thus becomes attainable in short iterations. Among its potential benefits are faster and more accurate disease diagnostics, the design of new drugs and therapies, and, in the longer term, the safe enhancement of the human species.