← back

About Mirror Life

Life Discovery Engine · port 8200 · LIGHT HOPE V5.3

The idea in one sentence

试活不同人生，更早认识自己。
Try many lives before reality gets expensive.

Why this exists

Most people find out what they love by accident — a job they happened to take, a field they happened to study, a decade that happened to pass. Mirror Life is built on a different premise: you can compress months of real experience into an hour of structured simulation, and the things you learn about yourself transfer.

Each world is a state machine + a knowledge graph + a 5-layer audit. You play through a scenario. The engine watches how you decide, not just what you decided. At the end, you get a process audit — a report on how you thought, separate from whether you happened to be right.

The currency is not entertainment. The currency is self-knowledge. The goal is to turn the ten career decisions you will make in the next ten years into ten informed ones, instead of ten accidents.

Five principles we refuse to break

  1. Thesis Before Capital. Every commitment must be preceded by a stated reason. No bet without a belief.
  2. Opportunity Cost Is Real. Every choice closes other doors. We show you what you gave up, not just what you took.
  3. Fear Narrows The Menu. When you are emotional, disciplined options are visually demoted — but never removed. You always see the whole menu.
  4. Loss ≠ Bad Decision. The audit measures process quality, not outcome. A well-reasoned loss beats a lucky win.
  5. Exit Is Always Available. You can leave any scenario at any moment without penalty. Engagement earned by obstruction is addiction, not experience.

What is in the engine

We do not hand-author content. Every scenario fact traces back to a citation — an annual letter, a peer-reviewed paper, a historical archive. The pipeline runs in five stages (antenna → knowledge graph → skills → assets → ingest) and every MHAP-1 asset is rejected at the validator if its sourceRefs array is empty.

This is what "no hallucinated content" means in practice. Editors cannot go off the rails because the engine refuses unsupported content. Look at any scenario node and you can trace it to a Buffett shareholder letter, a Kahneman paper, an SPIKES protocol citation.

Ecosystem

m200 is part of the LIGHT HOPE ecosystem of 350+ projects built around one question: what would compounding human development look like if we designed for it explicitly? Neighbors include m510 (Math Lighthouse), m595 (COMPASS rescue game — the pattern m200 generalizes), m554 (Cancer Truth Anchor), m8101 (AI OS orchestrator), and m8001-m8005 (governance and standards).

Roadmap

Contact

This is an in-development research product. Users of Round 3 should expect rough edges; the engine is real but the surface is minimal. Feedback that cites specific scenario nodes or audit scores is most useful.