For years the rule with AI chat was: keep threads small and single-focused. One chat per task. Finish, close, open a fresh one. Long threads rot. I learned that one the hard way in April, when I mined a month of my own sessions and found that past a hundred turns the answers stopped being right.
With Fable, I stopped.
One thread has been open for days now. I asked it what it was best at, and it said taking in a tremendous amount of information and not losing track of any of it. So I opened a single terminal window and we worked on five projects at the same time. It didn’t lose the thread once. It knew which piece belonged to which project and what we needed to do next.
When we got to 90% of the context window I asked whether to compact or start over. It said everything was documented, go to a new thread, here are three sentences to hand it. We moved threads once. Nothing important got lost.
I asked it why this works. The answer taught me something about my own system.
We moved the state out of the chat. The task board, the planning files, the scope docs, the PR descriptions, the memory files - that’s where the knowledge lives now. The thread is a control loop, not a filing cabinet. When compaction runs, it loses narrative, not decisions, because anything decision-shaped gets written to a durable file the moment it’s made. I call the habit OHIO. Only Handle It Once. Touch a decision, bank it, move on.
And the old rule didn’t die. It moved. Every piece of deep work still gets a fresh, single-purpose context. It’s just an agent now instead of a tab. The perpetual thread dispatches the work, reviews what comes back, and keeps the cross-project picture, which is the one thing that actually benefits from continuity. We industrialized the one-chat-per-task habit instead of abandoning it.
There are real risks. Compaction is silent, so the banking discipline can’t relax. If something only lives in the chat, it’s at risk. Long threads breed confident but stale beliefs, so the standing rule is check the ground truth before acting. And an old thread accumulates its own framings. For genuinely fresh strategic thinking, a clean room still beats the long conversation. That’s also why the agent that reviews a piece of work is never the agent that wrote it.
The thread never ends because nothing important lives in it.