Mastery: connect the pieces

1. Idea — Chat models do not remember yesterday by themselves. Memory is an application system for writing, storing, retrieving, and forgetting state across turns and ses…
2. Mechanism — Extract candidate memories, validate/consent, store with metadata, retrieve relevant items into the prompt, update or expire them, and keep audit trails.…
3. Controls — write policy, retention TTL, retrieval k, consent flags, memory scopes (user/session/org), forget API
4. Tradeoff — More durable memory feels personal and raises privacy/stale-advice risk. Aggressive forgetting is safer and more annoying. Automatic writes scale and can store …
5. Failures — False memory write; Stale preference
6. Metrics — useful-memory precision, contradiction rate, forget latency, consent coverage, tokens/turn from memory
7. Ship rule — Write only explicit preferences and mastery skills with consent; retrieve top 3 by relevance; TTL 180 days; ship forget in UI.

Neighboring layers (retrieval, serving, policy, human review) still own what conversation memory cannot on the personal tutoring assistant. Do not ask this chapter’s dial to replace permissions or source truth.

Mastery is transfer: take conversation memory into a second scenario that is not the personal tutoring assistant and rebuild the checklist without copying sentences. If you can only recite this chapter’s examples for conversation memory, you have memorized a story, not a model.

Also pin one numeric memory from this conversation memory chapter: If 40 memories are injected every turn but only 3 affect the hint, you spent tokens and risked contradictions—measure useful-memory rate, not memory count. That number is not decoration; it is a template for how claims about conversation memory on the personal tutoring assistant should look in design docs. Scoped specifically to conversation memory / personal tutoring assistant / mastery-connect.

People confuse conversation memory with neighboring buzzwords when debugging the personal tutoring assistant. Before changing prompts, ask whether the broken stage was evidence gathering, the conversation memory judgment itself, validation, or the product action. Fixing the wrong stage creates folklore (“we tried conversation memory and it failed”) that blocks the next team on the personal tutoring assistant. Scoped specifically to conversation memory / personal tutoring assistant / mastery-connect.

Write a five-line artifact for this page: goal, inputs, check, owner, stop rule. Invent one fluent failure that the check would catch. Keep details specific to conversation memory rather than generic AI advice.