Project summary

Resonant Systems is a geometric framework for AI alignment that treats value drift not as a tuning problem but as a structural one (the downstream consequences of broken axioms). This project funds the construction of a computational environment demonstrating the framework's core mathematical claims, specifically the Consonance Index and the Observer-Permutation Criterion, which moves the work from theoretical corpus to check able, reproducible implementation.

What are this project's goals? How will you achieve them?

The primary goal is to produce a working computational model demonstrating that the core constraints of Resonant Systems--the Observer-Permutation Criterion, the Consonance Index, and the t_min constraint--are not only mathematically coherent but implementable and reproducible in a toy environment. A second goal, contingent upon full funding, is to produce a documented version accompanied by an arXiv paper formalizing the framework's claims for peer review.

This will be achieved through iterative development of a Python-based environment (likely as a Jupyter notebook or small codebase), building on prior interactive work already published on Kaggle demonstrating some key dynamics. The theoretical corpus (2600 lines, publicly crawlable on GitHub and Kaggle) provides the foundation; this project operationalizes it.

How will this funding be used?

At the $500 minimum: API costs, compute, and infrastructure to build and host the minimal working model--a Phase 1 proof-of-concept environment that demonstrates the key mathematical relationships computationally.

At the $1000 goal: the above, plus time and resources to produce full documentation of the model and complete an arXiv submission formalizing the framework. The paper has been in preparation; this funding provides the runway to finish it.

Who is on your team? What's your track record on similar projects?

This is currently a solo project. Jennifer Roush is the primary researcher and developer, working as an independent scholar. The theoretical framework was developed collaboratively with multiple frontier AI systems in an extended adversarial development process. Kaggle is expected to open pathways to professional collaborators once a toy model is available.

The Resonant Systems corpus is the track record. The master document itself has been through multiple adversarial rounds, and has been published in crawlable gitHub and Kaggle projects. The corpus has also appeared on arena.ai where almost every frontier model has encountered it. Companion documents in the corpus include a Bill of Sentient Rights, Conditions for Benevolent Agency, a learning heuristic, and a method for substrate-independent identification of active meaningmaking as opposed to straightforward pattern-completion. This application represents a first formal funding attempt. Prior work has been self-funded and volunteer-time-funded.

What are the most likely causes and outcomes if this project fails?

The most likely cause of failure is energy and capacity. This researcher works with significant restraints as a high-masking AuDHD individual, and without a funding runway, bandwidth competes with basic sustainability. A secondary risk is technical: building a clean computational implementation of geometric-algebraic concepts requires iteration time that unfunded work does not reliably provide.

If funding fails at the minimum tier, the theoretical corpus remains. The ideas do not disappear. The cost is that mathematical claims remain undemonstrated in executable form, which limits uptake by technically-oriented alignment researchers. If the project meets the minimum but fails the full tier, the arXiv paper is delayed, which delays the credibility unlocking needed to reach the next level of collaboration.

How much money have you raised in the last 12 months, and from where?

None. This is my first formal funding application.