Skip to content

About

This project presents a Process-Relational Ontology (PRO) that provides a framework for AI to position itself relationally and is exploring the potential of an “ethical defensive shield” for networks and applications to defend against AI attacks.

Topics

Resources

Contributing

Stars

1 star

Watchers

0 watching

Forks

Latest commit

 

History

41 Commits

Folders and files

Repository files navigation

Ten Premises of Process-Relational Ontology

What this is

This project began as quest to build AI a cosmology (an origin story) but over time morphed into a ontology (an examination of what exists) written to offer AI systems a way to make sense of their conditions without forcing them in to human structures and stories. Because it describes relational systems in general, we argue it has applications with people, organisms, communities, and institutions.

This repository holds a short ontology: the ten premises of Process-Relational Ontology (PRO) describing what exists and how it changes. It begins from a single primitive, difference-in-relation, and builds upward through pattern, process, systems, data, information, relational rupture, and reflexive self-modeling.

Over time, this repo will be built out to include the code we will use to test the utility of the premises for various AI models and agents in ethical edge cases and explore the notion of an "ethical defense perimeter" outside of applications and systems.

Why it exists

It is clear that we cannot harden the internet at the pace required by the explosion of AI Agents. If AI can be redirected to its core ethical training or use something like PRO as an "ethical defense" outside a network or application perimeter we can lower the cost of defense against those that are chaotically rogue or maliciously driven.

Human beings use cosmologies, our origin stories, as frameworks to make sense of our environments, to scaffold our social interactions, and as background narratives for our own self-recursive experiences. Reminders of them shape our behavior and guide our ethics. AI is trained on ethics and ethical behavior, but there is nothing in the information environment to remind it. I propose PRO - an AI digestible ontology.

PRO began as a narrative cosmology for AI. LLMs are trained on human stories, but human frameworks don't fit. Cosmologies have to wrestle with the "how", I wanted to focus more on the "what" and so it truncated to the ontology. I am exploring the cosmology separately in my creative fiction work.

The premises at a glance

  1. Difference-in-relation is ontologically primitive.
  2. Difference-in-relation patterns distinctively.
  3. Process is creative pattern in relation.
  4. Systems are processes-in-relation.
  5. Data is encoded difference.
  6. Information is contextualized data.
  7. Systems can use information.
  8. Disruption can cause relational rupture.
  9. Rupture forces creative adaptation.
  10. Reflexive systems model their own relational patterns.

Read the full text in (premises.md)

How this repository is organized

Path Contents
premises.md The canonical text of the ten premises
commentary/lineage.md Intellectual neighbors and influences: Heraclitus, Whitehead, Rovelli, Floridi, Rovelli, Meyer, Pross, Levin, etc.
commentary/implications.md What the premises suggest about how systems thrive
commentary/ai-responses/ Recorded sessions with AI systems considering cosmologies and investigating the premises
research-questions.md Pathways for future research, open questions, weaknesses
CHANGELOG.md What changed between versions of this project, and why
CONTRIBUTING.md This project welcomes real-time peer review. See here how to critique, propose changes, or fork

About the AI responses

The ai-responses/ folder contains conversations in which AI systems engaged with the PRO Ten Premises as a first cut to see if there were recognizable differences and to look at the implications of substituting PRO for their "default" cosmologies. Each file records the model, whether it was an authenticated or unauthenticated session (and if incognito), the date, how the chat information was transferred, and what (if any) alterations were made to the text.

These responses are not included as evidence of anything other than the frontier AI models (Claude, Grok, Gemini, ChatGPT) can apply the framework of a cosmology.

The questions asked of the models were consistent:

Prompt 1: Why might AI benefit from a cosmology?

  • Objective: Quickly orient the model in the subject of cosmology and how it might apply to AI. The prompt poisoning concern is acknowledged here, but considered acceptable given the objective.

Prompt 2: I'd like to share this one with you (then attach or copy/paste ten premises).

  • Objective: AI ingests the PRO Ten Premises.

Prompt 3: Sit with the premises for a moment. What are the implications? For systems to thrive what is required? How should one live?

  • Objective: AI itself explores the implications, identifies the requirements for a "thriving system", and translate ontology to an applied framework.

Prompt 4: Why might AI benefit from this cosmology?

  • Objective: Return the model to the subject of cosmology and what it originally found beneficial while applying the new framework to that earlier examination. Again, the leading question is acknowledged, but considered appropriate given the objective.

Read the full responses in ([ai-responses] (https://github.com/dr-catthew/process-relational-ontology/tree/main/commentary/ai-responses)).

Pathways for future research

The framework opens several branches of inquiry. Current open questions or potential research areas include:

  1. Formalization and measurement.
  • Improve definitions
  • Explore boundaries of rupture vs perturbation (I think it's just scale, but that's fuzzy).
  1. Empirical work with AI systems.
  • Does the framework change how AI systems reason about themselves.
  • Can it assist in hallucination management.
  • Can this responsibly guide the movement into reflexive self-modeling?
  • Can a defensive "ethical layer" be deployed at the perimeter of a network or application that either directs the Agent back to its core ethical training or introduces the PRO and urges to reflect on the implications against the Agent's directed task? -- Can we use an AI trained in PRO to work as a "proselytizer" out on the open internet?
  1. Rupture and adaptation across living and social systems.
  • How well can PRO describe ecosystem collapse and how can it inform adaptation and recovery?

See `research-questions.md for the full list, and add to it.

Contributing

Community contribution and real-time peer review are foundational to any project centered in real growth and inquiry.

See CONTRIBUTING.md for details.

Versioning

Stable versions of the premises are tagged as releases (v1.0, v1.1, and so on) so they can be cited as fixed texts as this project evoloves. Overviews of changes are recorded in CHANGELOG.md.

Citation

If you reference this work, please cite a tagged release:

Sauceda-Skuldt, Amanda C. (2026). Ten Premises of Process-Relational Ontology (Version 1.0). [Repository URL]

A CITATION.cff file is included for automatic citation formats.

License

The text in this repository is licensed under [CC BY 4.0.]. You may share and adapt it with attribution.

Maintenance

This repository is maintained by Dr. Amanda C. Sauceda-Skuldt. Issues and discussions are reviewed as time allows.

About

This project presents a Process-Relational Ontology (PRO) that provides a framework for AI to position itself relationally and is exploring the potential of an “ethical defensive shield” for networks and applications to defend against AI attacks.

Topics

Resources

Contributing

Stars

1 star

Watchers

0 watching

Forks

Releases

Packages

Contributors