Skip to content

metanthropic/website

Repository files navigation

Metanthropic: Architecting the Physics of AGI

Metanthropic is a frontier research institution dedicated to building verifiable, interpretable, and safe Artificial General Intelligence. We move beyond probabilistic scaling to engineer deterministic systems where safety and reasoning are verifiable, intrinsic properties of intelligence.

⚛️ Our Thesis: The Glass Box Paradigm

Standard AI development relies on "Black Box" scaling, where controllability decreases as capability grows. Metanthropic rejects this alchemy. We operate on the conviction that to trust a mind, you must be able to read its thoughts.

  • Mechanistic Interpretability: We map neural circuitry to identify the origin of deception and reasoning failures.
  • Constitutional Design: Safety axioms are embedded directly into the pre-training objective, penalizing harmful trajectories at the atomic level.
  • Verifiable Reasoning: We build "System 2" engines that expose their chain of thought for human verification.

🛠 Tech Stack & Architecture

Built with a "Mission Control" aesthetic using cutting-edge frontend technologies:

  • Framework: Next.js 15 (App Router)
  • Content Engine: MDX (Markdown for Research) with dynamic slug routing.
  • Styling: Tailwind CSS with custom scientific color palettes.
  • Animations: Framer Motion for data-dense UI transitions.
  • Icons: Lucide React for technical and scientific symbology.

📂 Repository Structure

  • `src/app/`: Institutional interface (Safety, Research, Support Hub).
  • `src/content/research/`: MDX-based research publications.
  • `src/components/ui/`: Reusable design system (InteractiveGrid, DecryptedText, SpotlightCard).
  • `src/lib/`: MDX utilities and data handlers.

🚀 Getting Started

```bash npm install npm run dev ```

📄 Contributing Research

To add a new research paper, simply add an `.mdx` file to `src/content/research/` with the appropriate frontmatter.

⚖️ License & Charter

Operating under the Metanthropic Charter: our primary fiduciary duty is to humanity. We commit to stop competing and start assisting any value-aligned project that comes close to building AGI before we do. EOF

Releases

No releases published

Packages

No packages published

Languages