Project summary

Over the past several years, while building engineering systems and working extensively with increasingly capable AI models, I repeatedly encountered the same limitation: the AI was improving much faster than my own ability to process information. During extended interactions, my attention, working memory, and decision-making increasingly became the limiting factors.

That observation led me to ask a simple question: if humans remain part of the AI control loop, why do we rarely model the human side of that loop with the same rigor that we apply to AI systems?

This question became the foundation of YXFD, a cybernetic framework that models a person as a bounded computational system rather than an ideal decision-maker.

The framework consists of four interacting components:

• Y – the biological hardware that determines the physical limits of cognition.

• X – incoming sensory information and environmental inputs.

• F – active cognitive processing, including working memory, competing mental processes, and attentional allocation.

• D – judgment and behavioral output.

Rather than assuming unlimited cognitive capacity, YXFD models cognition as constrained by physiology, information overload, unresolved cognitive loops, and continual prediction about future events.

I believe this perspective addresses an underexplored aspect of AI alignment. Much of today's discussion focuses on aligning AI with human values. My work instead asks how humans themselves can remain reliable decision-makers while interacting with increasingly capable AI systems. If humans are expected to supervise AI, then human cognitive limitations should be treated as part of the alignment problem.

This project is not intended to present YXFD as a finished scientific theory. Instead, the objective is to transform several years of independent research into a rigorous manuscript that clearly defines the framework, compares it with existing work, and invites critical discussion, refinement, and empirical testing.

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

The primary goal is to transform YXFD from a personal research framework into a formal research contribution that other researchers can evaluate, criticize, and extend.

Over six months, I plan to:

• complete the first full manuscript describing the framework;

• review relevant literature in cybernetics, neuroscience, cognitive psychology, human-computer interaction, and AI alignment;

• Refine the formal definitions and notation used throughout the model;

• develop diagrams and computational representations of the framework;

• compare YXFD with existing approaches and identify both similarities and differences;

• seek feedback from researchers across multiple disciplines; and

• Prepare the work for public release as an open-access manuscript.

I am not claiming that YXFD is already validated. My objective is to present the framework with enough clarity that its assumptions, predictions, strengths, and limitations can be examined through scientific discussion and future empirical research.

How will this funding be used?

The funding will primarily provide dedicated research time.

For several years, I have developed this work alongside significant financial constraints. Funding would allow me to spend six months focused exclusively on transforming scattered notes, models, and prototypes into a coherent research manuscript.

Specifically, the funding will support:

• six months of full-time research and writing;

• access to academic literature and research databases;

• professional editing and manuscript preparation;

• production of technical diagrams and figures;

• computational resources for exploring future AI implementations of the framework; and

• Publication or conference-related expenses.

The first milestone is not a commercial product.

It is a research publication that clearly explains the framework and makes it available for critical evaluation by the broader scientific community.

At the end of the project, I intend to release:

• a complete research manuscript;

• formal definitions of the YXFD framework;

• system diagrams and computational representations;

• a comparative review of related literature; and

• An openly accessible version of the research for public discussion.
  

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

I am currently the sole researcher working on YXFD.

My background is in computer engineering, and for several years I have worked independently on projects involving engineering, systems thinking, and AI. Much of this work has been self-funded, giving me the freedom to develop ideas over long periods without institutional support.

YXFD is the result of years of iterative development rather than a recently conceived idea. During this time, I have produced working documentation, conceptual models, experimental AI-assisted cognitive workflows, and an early conversational prototype ("Mirror") that applies parts of the framework during structured reflection. These prototypes have helped me refine the concepts and identify areas requiring further development.

One purpose of this funding is to move the project from independent research into broader scientific discussion. As the framework matures, I hope to collaborate with researchers in neuroscience, cognitive science, AI, cybernetics, and human-computer interaction to test, challenge, and improve the model.

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

The greatest risk is not that the framework turns out to be incorrect. Scientific progress often requires revising or rejecting ideas when confronted with evidence.

The more immediate risk is that the project remains unfinished due to limited time and financial resources. As an independent researcher, balancing research with everyday financial obligations has significantly slowed progress.

Another possibility is that parts of the framework overlap substantially with existing research or require significant revision after comparison with the scientific literature. If that occurs, I would consider it a successful outcome of the research process rather than a failure.

Even if YXFD is not ultimately validated as a complete framework, I believe its central question—how to formally model the cognitive limitations of humans within AI-assisted decision-making—will remain valuable and may contribute useful ideas for future work in human-AI collaboration.

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

I have not received external funding for this project during the past twelve months.

Development has been supported entirely through my own time and personal resources. My priority has been to refine the underlying ideas before seeking financial support or institutional partnerships.

This application represents my first effort to secure funding dedicated specifically to transforming the research into a formal manuscript that can be openly evaluated by the broader research community.

Future Applications

Although this project focuses on formalizing the theoretical framework, one of its long-term motivations is to use YXFD as the foundation for AI systems that act as cognitive support tools. Rather than replacing human reasoning, such systems would help users recognize patterns of cognitive overload, identify recurring background cognitive loops, and allocate their limited attention more effectively during complex decision-making.

Developing such an AI system is outside the scope of this proposal. The goal of the current project is to establish the theoretical foundation that could enable future research and implementation.

Expected Deliverables (within 6 months)

• Complete research manuscript describing the YXFD framework.

• Formal notation and system definitions.

• Technical diagrams of the computational model.

• Comparative review of related literature.

• Open-access publication of the manuscript.

• Public repository containing the manuscript, diagrams, and supporting documentation.