---
title: "Memory.wiki Development Journey & Vision"
url: https://memory.wiki/MLA3xnP9
updated: 2026-05-20T12:37:51.841Z
source: "auto-synthesis"
---
# Memory.wiki Development Journey & Vision

> These documents capture the evolution of memory.wiki from initial concept to business plan, showing how it aims to solve AI knowledge delivery by making user knowledge accessible to any AI through URLs.

## Key claims

- [EXTRACTED] The core problem isn't AI memory but knowledge delivery: "매일 수백만 명이 ChatGPT, Claude, Gemini, Cursor에 자신의 사고를 쏟아붓니다... 그리고 탭을 닫습니다. 그 답변은 사라집니다" [doc-1]
- [EXTRACTED] Memory.wiki operates on a single architectural principle: "memory.wiki URL은 어떤 AI든 사용할 수 있는 API입니다" [doc-1]
- [EXTRACTED] The system uses a three-tier architecture: "3-Tier Architecture (수집/소화/활용)" with capture, digestion, and utilization phases [doc-3]
- [INFERRED] The project underwent significant rebranding and repositioning between versions, transitioning from "mdfy" to "memory.wiki" while maintaining the core URL-as-API concept [doc-3, doc-5]
- [EXTRACTED] The business model positions itself as "LLM 서비스의 기본 지원 메모리 레이어" (fundamental memory layer for LLM services) [doc-4]
- [AMBIGUOUS] The competitive differentiation combines user authorship with AI organization, but the exact boundaries between automation and user control remain unclear across documents [doc-1, doc-4]

## Cross-references

- **URL Architecture**: Both doc-1 and doc-5 emphasize URL-native design, with doc-1 describing `memory.wiki/<id>`, `memory.wiki/b/<id>`, and `memory.wiki/hub/<slug>` patterns that doc-5 calls "three tiers"
- **AI Integration**: The "어떤 AI든" (any AI) principle appears consistently across doc-1, doc-3, and doc-4, positioning the platform as AI-agnostic
- **Product Evolution**: Doc-4 and doc-5 show the transition from "mdfy" branding to "memory.wiki", with doc-3 representing the final business plan incorporating this change

## Open questions / gaps

- How does the automatic AI digestion work technically, and what level of user control exists over the organization process?
- What are the specific authentication and privacy mechanisms for knowledge sharing across different AI platforms?
- How does the 16-week launch timeline (mentioned in doc-3) align with the technical complexity of cross-AI integration?
- What is the actual user adoption strategy beyond the "viral badge loop" mentioned in doc-5?

## Provenance

- [doc-1]: Foundational manifesto explaining the core problem and solution concept for memory.wiki
- [doc-2]: Brief planning note confirming the v7 direction and feature retention decisions  
- [doc-3]: Comprehensive v7 business plan with architecture, roadmap, and strategic positioning
- [doc-4]: Simplified explanation of the memory.wiki concept and competitive differentiation
- [doc-5]: Technical product reference for the earlier "mdfy" version, showing product evolution