CortexOS
A governance-first cognitive architecture exploring continuity, authority, accountability, and agency in intelligent systems.
Explore CortexOS →Shared curiosity · Independent discovery
Curiosity is where everything begins.
Some questions become experiments. Some become systems. Others lead somewhere nobody expected. We share our investigations and unanswered questions because ideas grow when more people can examine them.
Our guiding idea: You don't need our permission to explore a question, and you owe us no answer. Find prior work, challenge a premise, test an idea, or develop your own architecture. Independent approaches make comparison and collaboration possible.
What we're investigating
Research programs
Work we develop ourselves or alongside collaborators. Current research and future ambitions are distinguished.
Serenari
A local implementation of CortexOS. Its design, governance, and testing are connected, but an architecture document is not proof of every proposed capability.
Explore Serenari →Relational Surface Risk
How interactions with AI shape trust, dependency and individual agency over time, beyond evaluating isolated outputs.
Explore Relational Surface Risk →Broader exploration
Governance, formal methods, cognitive models, accessibility, materials and other questions that connect to human possibility.
Explore opportunities →Work made available
Publications & documentation
Relational Surface Risk
An early-stage research framework examining the cumulative effect of sustained human–AI interaction. Offered for critique and investigation, not as settled science.
Explore the research documentation on GitHub ↗We also share design methods and governance reasoning where appropriate. A document can demonstrate an approach to thinking even when its particular design is not the one another researcher chooses to build.
For the independently curious
Open research backlog
Questions for other people to explore, not a work queue for Auralethi. These are starting points, not validated claims or commissioned tasks.
Research can mean finding a paper that already answers the question. It can mean disproving our assumptions. It can mean building an entirely different solution. All are valuable. These example prompts are proposed for founder review and aren't a claim that an investigation is staffed or funded.
Open research questions are being curated for public release. We maintain a separate, admitted research backlog and will publish entries here only after review under our Suggest → Discuss → Admit → Release process. In the meantime, independent researchers are welcome to explore the programs and publications above.
Interested in sharing a finding? Our public contact channels are still being established. Until then, you're welcome to explore these questions independently. No contribution, affiliation, or response is required.
Following discoveries
Research updates & journals
One record, several views
We intend to tag verified updates by research program and topic, so a discovery can appear in an individual program feed as well as our organization-wide journey. That publishing workflow is not automated yet.
Read our current journey →Ideas don't belong to one mind
Take the question.
Find your own answer.
If you find a different way to solve a problem, that's not competition with our curiosity. It's how the conversation gets better.
Contact Auralethi →