Project Description:

What I've Built:

I've developed an AI-powered analysis framework that maps structural dependencies in complex systems by preserving contradictions rather than collapsing them into simplified narratives.

The system works by:

Representing dependencies as multi-layered graphs

Tracking "pressure" from unresolved contradictions

Refusing to synthesize when synthesis would destroy information

Identifying non-fungible constraints (what can't be substituted/parallelized)

I've applied this to two domains:

1. Aerospace Industrial Base

Mapped organizational, technological, material, and market dependencies

Identified single-point failures and non-diversifiable constraints

Cross-validated against DoD and GAO industrial base reports

2. China Semiconductor System

Seven-layer fracture point analysis (lithography, EDA, equipment, materials, etc.)

Identified where unlimited investment still hits hard constraints

Distinguishes delays (solvable with resources) from structural impossibilities (physics/integration bottlenecks)

What Makes This Different:

Standard analysis asks: "What are the risks?"

This framework asks: "What cannot be changed regardless of resources?"

That difference matters for:

Strategic intelligence (knowing where systems actually break)

AI alignment (systems that preserve honest uncertainty)

Policy decisions (understanding genuine constraints vs solvable problems)

What I Need Funding For:

Currently working from my phone in Iqaluit, Nunavut. Need resources to publish properly:

$400 - Used laptop for faster writing/diagram creation

$200 - Professional attire for validation calls with experts

$1,500 - Living expenses to work full-time for 3 weeks instead of other income

$400 - Buffer for publication needs (software, unexpected costs)

Total: $2,500

Deliverables (by March 22, 2025):

Published case study: "Structural Fracture Points in China's Semiconductor System"

20 pages with dependency diagrams

Published on ArXiv, LessWrong, GitHub, personal site

Methodology documentation: Full explanation of constraint-preserving framework

How to replicate the analysis

Technical implementation details

Validation methodology

Expert validation: Feedback from at least 2 domain experts in semiconductor/tech policy

Documented in published report

Corrections/refinements incorporated

Open source: All analysis tools and methods made available for others to use

Why This Matters:

This work sits at the intersection of:

AI Safety: Demonstrates that LLMs can be made to preserve contradictions instead of hallucinating synthesis

Geopolitical Analysis: Provides framework for understanding structural constraints in tech competition

Complex Systems Research: Offers new methodology for analyzing systems where oversimplification is dangerous

Verification:

I can provide:

Draft sections of the semiconductor analysis

Demonstration of the methodology

References from anyone who wants to validate the work first

I'm happy to do a call to discuss the project with potential funders.

Budget Breakdown:

Item

Cost

Justification

Used laptop

$400

Currently working from phone; need proper writing/diagram tools

Professional attire

$200

For video calls with expert validators and potential stakeholders

Living expenses (3 weeks)

$1,500

To work full-time instead of other income ($500/week)

Publication buffer

$400

Software, diagram tools, unexpected costs

Total

$2,500