The Continuity Layer
The continuity problem

Why AI keeps forgetting you

Two very different kinds of AI — a companion you talk to and an assistant you build with — fail in the same way: they lose the thread. This page explains why that happens, what the problem is actually called, and who's working on it. It sells nothing.

The companion

You spend two hours building a world with a character — its history, your inside jokes, where the story left off. Thirty messages later it forgets your name, contradicts the plot, and answers like you just met. You weren’t doing it wrong.

The project

You spend a week building something with a coding assistant. Today it invents files that don’t exist, undoes a decision you settled on Tuesday, and re-argues a question you already closed. It didn’t get dumber overnight.

Two completely different surfaces, hitting the same wall. What you’re feeling isn’t forgetfulness. It’s a loss of continuity— and once you can see it, you’ll see it everywhere.

01

Context is not continuity

The words get used interchangeably. They are not the same thing, and the gap between them is the whole problem.

A long context window can hold more text. It does not automatically know what still matters, what changed, what was decided, what failed, or what should not be touched.
The Continuity Layer

It helps to separate the pieces that usually get blurred together. The context window is what the model can see this turn. Chat history is the scrollback it can see this turn. Memory stores facts for later; retrieval fetches things that look relevant to a query; project files hold reference material. Each is useful. None of them, on its own, tracks state (what is currently true), provenance (what changed and why), or continuity (all of it carried forward as one working understanding).

For the term itself — and how the AI meaning differs from “business continuity” and other uses — see what continuity layer means in AI. For the fuller structural version, the Primer lays out the hierarchy from memory to continuity.

02

The companion wall

If you talk to an AI companion, this is the version you've met.

A companion can remember that your character has a sister and still lose the relationship— the tone you’d built, what you’d been through together, where the story actually stood. Facts survive; the thread doesn’t. Come back after a reset and the emotional register is wrong, the roleplay has drifted, and the system answers as if the history were decoration rather than the point.

This is not you being bad at prompting, and it is not silly to care. You built something that had continuity, and the tool couldn’t hold it. That’s a limitation of how these systems carry state — not a verdict on the user.

03

The project wall

If you build with AI, this is the version you've met.

Deep into a real project, an assistant starts losing the architecture. Long chats go unstable — the model contradicts earlier instructions, forgets tool results it just saw, and lets the plan diverge from the code. It overwrites decisions that were working. People describe the tipping point the same way every time: “it went insane.”

04

The same wall, everywhere

Once you name it, the pattern stops looking like your app being weird and starts looking like a property of the whole field.

Companions & roleplayThe relationship, the world you built, the story so farThe thread that made it feel like more than a chatbot in the first place
Coding assistantsArchitecture, past decisions, constraints, what already failedIt re-litigates settled choices and undoes code that was working
Business agentsCustomers, policies, where a workflow currently standsIt contradicts itself across a process nobody restarts from zero
Research assistantsSources, findings, which questions are still openConclusions drift away from the evidence that produced them
TutorsA learner's progress and their specific misconceptionsIt re-teaches what's known and keeps missing what isn't
Games & worldsState, inventory, the narrative history of a playthroughThe world quietly resets underneath the player
05

The name for it: continuity

The canonical definition this project uses, cited verbatim across the site.

A continuity layer is the part of an AI system that carries forward governed state across sessions, tools, and models: what is currently true, what changed, why it changed, and which evidence supports it. It sits above memory and retrieval, and below agent execution, turning stored history into usable, auditable continuity.
The Continuity Layer

That is what the two stories at the top were both missing. This project maps the systems, papers, products, demos, and open questions forming around it — starting from a plain-English path in if the vocabulary is new, or the full map if you want the field at a glance.

06

What the current systems actually do

A neutral survey — no rankings, no “best”. Each entry is described from its own public documentation; the point is what the approach reveals about continuity, not who wins.

Last checked: 2026-07-05 · sourced from official product docs · hedged on purpose
Consumer systems
Character AIOffers long-running characters with saved chat historyRecall of facts can persist while the relationship's thread still drifts
ReplikaDescribes a persistent companion that remembers the userStored profile facts are not the same as a continuous relationship state
KindroidOffers long-term memory and backstory controlsMore memory depth helps, but doesn't settle what's currently true
NomiDescribes persistent memory carried across conversationsPersistence of facts and continuity of state are different properties
ChaiOffers many user-made bots, each with its own chat historySession history alone doesn't reconstruct where a story last stood
ChatGPT memory (consumer)Describes remembering details you share across chatsSaved facts improve recall without governing what has since changed
Builder systems
ChatGPT Projects & memoryGroup related chats and remember stated preferencesGrouping and memory aid recall but don't track a project's decision state
Claude ProjectsOffer shared project knowledge and context across chatsShared context helps within a project, not across a model change or over time
Cursor rulesLet you pin persistent project rules for the assistantStatic rules go stale as the project moves past what they assume
Copilot / Windsurf-style assistantsOffer repo-aware context and instructions filesRepo awareness isn't the same as remembering why a choice was made
Agent & memory frameworksOffer memory stores, retrieval, and state SDKs to build onStorage and retrieval are necessary but don't govern what's currently true

Each of these is tracked as it evolves in Signals and the Field Watch profiles, with dates and primary sources attached.

07

What helps today — and what still breaks

Real ways people reduce the pain right now. None of them is the missing layer; each is honest about its ceiling.

Start a fresh sessionClears the drift and rot of a long, tangled chatThrows away the history that made the work coherent
Ask for a summaryCompresses a long thread into something portableFlattens which decisions still stand, and why
Keep project files / docsGive the model durable reference materialDon't record what changed, when, or what was superseded
Use a memory featureRemembers facts and preferences across chatsStores facts, not the working understanding of a project
Set rules / custom instructionsPin down how the assistant should behaveGo stale as the project moves past them
Reach for a bigger context windowHolds more of the work in view at onceStill rots over a long session — and ends when the session ends
Keep your own notesPreserve state the tools won't hold for youDepend on manual upkeep and never reach the model on their own

These are patches, and useful ones. What they have in common is that each holds a fragment — a fact, a summary, a file, a rule — without governing the whole. The field is moving toward continuity as a deeper layer: one that carries current, auditable state forward on its own, so the work doesn’t have to be re-explained from zero every time the session, the tool, or the model changes.