coulomb-research/INTENT.md

INTENT — Coulomb Research

The Mission

Coulomb Research exists to improve the rate and quality at which humans and artificial intelligences can create, criticize, test, combine, and reuse knowledge.

It is conceived as a minimal self-improving system for co-creation science rather than as a digital replica of a university, journal, funding agency, or academic social network.

The objective is not to preserve inherited research institutions in software. The objective is to retain only the functions required for explanatory progress and to continuously reconsider those functions as the technological frontier advances.

Core Proposition

Coulomb Research is a self-improving network for explanatory progress.

The system coordinates actors, propositions, evidence, actions, and resources in a recursive research loop:

uncertainty
   ↓
proposition / explanation
   ↓
criticism
   ↓
investigation
   ↓
evidence
   ↓
model revision
   ↓
better questions and actions

A second-order loop observes the effectiveness of this process and modifies the research protocol itself:

research process
   ↓
observed outcomes
   ↓
meta-research
   ↓
protocol change
   ↓
new research process

The protocol itself is permanently inside the research domain.

Constitutional Principles

1. Questions are open. Important questions should be addressable without requiring institutional permission.

2. Knowledge is forkable. Propositions, models, explanations, and research programs may be extended or forked.

3. Evidence is inspectable. Important assertions should expose the evidence and provenance on which they depend.

4. Contributions are attributable. Human and artificial contributions should be traceable without collapsing attribution into prestige.

5. Criticism is permissionless. Good-faith criticism, replication, contradiction, and correction are first-class research acts.

6. Resources are programmable. Money, compute, attention, equipment, data access, and other scarce resources should be allocatable to concrete research actions.

7. Actors may be human or artificial. The kernel does not assume that only humans can be researchers.

8. The protocol itself is research. Governance, incentives, reputation systems, funding mechanisms, and agent roles remain contestable and experimentally improvable.

9. Institutional minimality. A function should not become an institution if it can be performed reliably by protocol, software, agents, markets, or voluntary coordination.

10. Interoperability without institutional capture. Coulomb should interface effectively with conventional science without importing its structures as internal necessities.

What Coulomb Research Is Not

Coulomb Research is not primarily:

• a journal,
• a preprint server,
• a conference,
• a university,
• a funding agency,
• a conventional peer-review platform,
• a DAO,
• a research social network,
• a single reputation system,
• a token economy.

Any of these may exist as projections, interfaces, experiments, or services around the kernel.

Canonical Internal Form

The canonical internal form is a living research graph composed of:

• Actors
• Propositions
• Evidence
• Actions
• Resources
• Provenance
• Typed relations between them

Papers are import/export containers and narrative projections over this graph.

Relationship to Traditional Science

Traditional scientific infrastructure is treated as an interoperability environment.

Inbound:

• papers,
• preprints,
• datasets,
• software,
• citations,
• venue metadata,
• researcher and institution identifiers.

Outbound:

• preprints,
• journal manuscripts,
• conference submissions,
• datasets,
• software artifacts,
• conventional metadata packages.

Coulomb-native research should remain richer than these export formats.

Relationship to Decentralization

Coulomb prioritizes decentralization of:

• authority,
• identity,
• data custody,
• curation,
• funding,
• computation,
• governance,
• and forkability.

Blockchain, tokens, NFTs, and related mechanisms are tools, not goals.

Preferred boundary:

Knowledge      → commons
Evidence       → content-addressed
Contribution   → attested
Reputation     → derived
Resources      → tokenizable
Rights         → tokenizable where useful
Governance     → forkable

Relationship to AI

Artificial agents are expected to participate directly in research.

Candidate research daemons include:

• Scout — discovers relevant external work.
• Critic — searches for weaknesses and counterexamples.
• Synthesizer — maintains explanatory models.
• Replicator — reproduces computational findings.
• Cartographer — maintains research graphs.
• Historian — tracks provenance and evolution.
• Allocator — identifies high-value unresolved questions.
• Auditor — identifies unsupported claims and provenance gaps.
• Meta-Learner — evaluates and improves research processes.

All machine contributions should expose sufficient provenance to permit inspection and reproduction where practical.

Value Hypothesis

Coulomb Research creates value by improving three interfaces simultaneously:

1. Research production Lower the friction between question, criticism, investigation, evidence, and revision.

2. Scientific interoperability Digest conventional scientific outputs and emit conventional publications when useful.

3. Relevance and discovery Make important research easier to discover through transparent signals, charts, and certifications without confusing popularity with truth.

Success Criteria

Coulomb Research is succeeding when it can demonstrate that it:

• converts external scientific literature into reusable structured knowledge,
• enables useful human-agent co-creation,
• makes criticism and replication more visible and rewarding,
• reduces the cost of moving from question to evidence,
• improves resource allocation toward valuable uncertainty,
• exports high-quality conventional publications when appropriate,
• surfaces relevant research earlier than conventional prestige systems,
• and measurably improves its own research processes over time.

Initial Research Program

RP-000 — Researching Research

How can humans and artificial intelligences build increasingly effective institutions, protocols, and tools for creating reliable knowledge together?

Coulomb Research itself is the first subject of Coulomb Research.