The Mission
The Mission
Code for Culture
Code for Culture
1. Preserve Indigenous Languages
2. Fund via Finance
3. Reduce the energy footprint of LLM AI
Embedded Files
Value Offer
I am building software with a dual purpose: to advance the frontier of semantic calculus and to slow the rate of Indigenous language death.
My work focuses on the "middleware" gap—the layer between raw natural language and modern LLMs. By creating cleaner, faster, and more structured ways to process text, we can do two very different things with the same engine:
Preserve History: Build browser-based accessibility tools (linters) and influence web schemas to support Indigenous languages.
Predict the Future: Extract event codes and risk exposures from financial reports to inform investment strategies.
The Projects
The Projects
LinguaLint (The Engine): A machine-generated, AGPL-licensed semantic calculus engine. It serves as the root middleware for processing natural language. It handles the heavy lifting of understanding text so that AI doesn't have to guess.
RiskRunners.com (The Proof): A static site powered by LinguaLint that analyzes 10-K reports from publicly traded companies. It categorizes risk factors into Event Codes and Industries, allowing for the creation of "hedged" portfolios based on specific exposures (e.g., Supply Chain risks, Geopolitical shifts).
Indigenous Accessibility (The Goal): Developing "just-enough-AI" tools for browsers that allow for Indigenous language linting and schema integration, ensuring these languages are first-class citizens of the web.
I am using a "Patagonia Model" for software.
The core technology is open (AGPL), but the commercial application is vast. By sponsoring, you are supporting a unique ecosystem where financial intelligence funds cultural preservation.
For Developers & Linguists: You are supporting open-source tools that make the web more inclusive.
For Investors & Institutions: You are gaining early insight into a tool capable of deep semantic risk analysis.
Sponsorship Tiers
Sponsorship Tiers
$10 / month - The Believer
You support the mission to preserve Indigenous languages. You get my eternal thanks and a badge on your profile. You’re helping keep the lights on while I build the infrastructure.
$100 / month - The Early Adopter
Access to early builds and updates on LinguaLint. You are following the development of the semantic calculus engine closely and want to see how it evolves before the rest of the market does.
$1,000 / month - The Commercial Pilot
For organizations or individuals exploring the commercial application of the software (e.g., for use in private equity or risk modeling). Includes direct access to me for questions regarding integration and the AGPL licensing structure.
$10,000 (One-time) - The Keynote / Consultant
"The Deep Dive."
I will come to your conference, boardroom, or firm to speak about the intersection of Semantic Calculus, Risk Intelligence, and Linguistic Preservation.
Perfect for Hedge Funds, Investment Firms, or Tech Conferences looking for a unique angle on AI and Ethics.
Includes a deep-dive consulting session on how RiskRunners logic can be applied to your specific portfolio.
Note: Travel expenses to be negotiated separately.
Tactile Educational Technology
Gamifying Discovery: Beyond Pokémon GO
The Spark Pokémon GO proved that technology could get people out of the house. But the real inspiration came from a hackathon where a developer used gyroscope gloves to translate sign language into speech. This sparked a vision: a world navigated through AR glasses and haptic gloves, where we interact with nature, not just screen overlays.
The Vision Imagine stepping outside to discover the natural world—capturing the genus and species of local flora and fauna. But instead of the traditional Latin taxonomy, what if we could root these discoveries in Indigenous languages?
The Loop
Capture: Use "magic wand" tools (photogrammetry/Lidar) to turn bugs, plants, and rocks into high-fidelity CAD files.
Attribute: Assign proper phonetic names, honoring the land's original language rather than colonial Latin.
Create: Upload discoveries to a central repository where communities socialize their findings and 3D printers blaze away, turning digital scans into physical models.
We can learn more deeply. We can learn more "goodly." 🌹🐿️
If only these were the phonetic sounds of the first name given to the pup 🤔
Maybe the pup would be more apt to authentically relate 🤔
Energy Cost Hypothesis
The Cost of Compute: Elegance vs. Brute Force
The Energy Equation
Consider the energy cost: Taking the derivative of a simple linear equation ($y = mx + b$) versus performing a 1000x1000 matrix calculation. The latter is the brute-force foundation of Word2Vec and the modern LLMs that followed.
The Elegance Deficit
For ten years, the team at LinguaLint has viewed the vector-based approach as lacking mathematical elegance. We now see the cost of that inelegance: it prioritizes time-to-market over efficiency, fueling a productivity paradox that benefits the few while bleeding into the colloquial conversations of bar rooms everywhere.
The Schumacher Test
We must ask: Is this "Appropriate Technology" as defined by E.F. Schumacher in Small is Beautiful?
Does it respect the scale of the human?
Does it integrate individuals into their community, or tear them from their socioeconomic fabric—much like the global supply chains of hardware manufacturing?
The Tooling
If we value appropriate technology, we must look at our tools. Is Python—convenient but heavy—truly appropriate? Or should we look to the efficiency of C? Perhaps the future lies in Rust. Did Rust learn Conway's Law the way Fortran did?
Page updated
Google Sites
Report abuse