For arXiv and research visitors

Disruption Joe Research

I am Joe Hernandez. These public repositories are my working proof of competence in AI-native epistemic research systems: research programs that make hypotheses explicit, expose proof burdens, keep failure conditions visible, and use AI agents as disciplined stewards rather than as hidden authors of certainty.

The claims below are not presented as a single finished theory. They are an open research stack: some results are theorem-grade, many are finite-witness or speculative, and the governance machinery is part of what is being tested.

What this page is

A public index for an AI-native research system.

My research work sits at the intersection of foundations, governance, and epistemology. The recurring question is whether AI-native systems can help us build research machinery that is more explicit, more adversarial, more reproducible, and more honest about what has and has not been earned.

Each repository has its own domain truth. The through-line is the method: claim ledgers, adversarial personas, absorber maps, tests, kill criteria, provenance records, and explicit status labels. The point is not to use AI to make research sound more confident. The point is to use AI to keep the research harder to fool.

Repository map

Five public research surfaces.

The order below starts with the most direct evidence of the AI-native epistemic machinery, then points into the domain programs where that machinery is being tested under pressure.

Meta-research / epistemic architecture

AI-Native Evolutionary Knowledge Systems

This repository asks what changes when the machinery that produces knowledge becomes an explicit design object. It studies whether AI-native systems can make reasoning roles, verification loops, incentive separation, adversarial review, memory, and claim governance modular enough to be designed, tested, copied, and evolved.

Open repository
Core question Can epistemic machinery itself become an engineering object?
Evidence source The adjacent research repos are treated as observations of recurring governance patterns, not as assumed truths.
Current burden Show residual capability beyond known social, institutional, and inference-time search explanations.
Foundations / record finality

Time as Finality

A claim-led formalization program around the idea that the past is what has become hard to undo. It asks whether observer-relative temporal order can be reconstructed from causal order and stabilization frontiers of accessible records. Its strongest discipline is the distinction between finite executable witnesses, general theorems, and open physical interpretation.

Open repository
Foundations / issuance

Temporal Issuance

A speculative but adversarially governed research program asking why reality remains capable of producing genuinely new structure. Its working hypothesis is that reality is a shared, distributed, observer-participatory process kept open by ongoing issuance, while temporal order and finality are downstream reconstructions from records.

Open repository
Physics / geometry pressure test

Observerse Research Program

This program began as Geometric Unity formalization and advanced into a broader study of Clifford-Rarita-Schwinger / chimeric-bundle observerse geometry. Its current public floor is a GU-independent structural law: this class of matter geometry is intrinsically vectorlike and cannot force its own chirality or generation count from inside.

Open repository
Institutions / contribution legitimacy

Architecture of Legitimacy

A public research program on legitimate contribution allocation in open projects: how contribution, review, credit, reward, contestability, retained record, and governance transition might cohere without assuming that open participation is automatically fair or that AI judgment is neutral.

Open repository
Suggested path

If you are deciding whether to take this seriously.

Start with the method and then sample the pressure tests. The goal is not to agree with every hypothesis. The goal is to see whether the research system keeps its own uncertainty visible while still producing useful work.

1. Read the meta-claim

Begin with AI-Native Evolutionary Knowledge Systems to see the explicit claim about epistemic machinery as an engineering object.

2. Check a hard domain

Use the Observerse/GU repo to see how bold conjectures are converted into falsifiable pressure, correction logs, and proof-grade status labels.

3. Inspect the governance

Look at Temporal Issuance and Architecture of Legitimacy for the steward, claim-ledger, contestability, and non-capture machinery.

Research contact

For critique, collaboration, or review.

I welcome serious objections, absorber mappings, replication attempts, and domain-specific correction. The cleanest entry point is usually a GitHub issue in the relevant repository.

View GitHub profile