Machine Memory · v0.1
Table of Contents
A reference manual for how machines learn, remember and forget. Chapters in grey are being written or drawn; subscribe and each one reaches you the day it is finished.
Finished chapters:
16 / 25
Words/chapter:
1.7K
Images/chapter:
2
Current word count:
27.4K
Est word count:
42.8K
Current image count:
36
- 1. Foundations
Before a machine can remember anything it has to turn the world into numbers. These chapters are about the numbers: what a neural network is, how text becomes a vector and why attention changed everything.
- 2. Memory
A model on its own remembers nothing between conversations. Everything that feels like memory is a system built around it: what gets written down, how it is found again, and when it is let go.
- What does it mean for a machine to remember?planned
- Retrieval.planned
- Extraction and summaries.planned
- Forgetting.planned
- 3. Hardware
Models are maths, but the maths runs on silicon. Where the numbers physically live decides how fast a model thinks and how much it can hold.
- How does a GPU work?in progress
- The memory hierarchy.planned
- Running a model on your laptop.planned
- 4. Protocols
Assistants talk to tools, files and each other through a handful of protocols. Knowing what goes over the wire makes them much less magical.
- How does MCP work?planned
- Tool calls.planned
- 5. Guides
Practical, shorter pieces: how to give the assistants you already use a memory that lasts, and when you do not need a tool at all.
- How Esment works.1.3K words
- A local MCP memory server.1.7K words
- Shared memory for Claude Code and Cursor.1.7K words
- CLAUDE.md or a memory server?1.6K words
- Persistent memory in Cursor.1.7K words
- One memory for ChatGPT and Claude.1.7K words
- Does Claude remember across chats?1.1K words
- Making ChatGPT remember.949 words
- ChatGPT memory full.897 words
- Long-term memory for agents.1.4K words
- AI with memory.1.2K words
- Claude Code context.1.2K words
New chapters, by email
The manual is written in the open, a chapter at a time. Leave your address and each new one lands in your inbox the day it is published. Nothing else, and one click to leave.
Want to start now? Read the first chapters