Durability
465 confirmed writes.0 violations observed.
How do we check that a confirmed piece of information does not silently disappear?
Lab
We study how humans and AI lose, find again and use context.
Our goal:
keep less, keep better, and bring back the right information at the right time.
It is a system we keep improving so that AI understands your context better.
Durability
465 confirmed writes.0 violations observed.
How do we check that a confirmed piece of information does not silently disappear?
False precedents
We study how to tell apart:
Non-writing
Our first synthetic non-writing protocol correctly classified 26 events out of 26.
Reality will be more complicated.
That is precisely what we want to observe next.
Import
Imported information stays identified as such and never automatically becomes authoritative.
It does not try to remember everything.
It tries to make your experience available again when it matters.