Featured explainerVideo
The cognitive layer for AI
A digital mind for every identity, human and agent.
8:42
Understand the cognitive layer from every angle.
One platform, two ways in: Learn how to get the most out of thinqOS, and explore the Science behind it: video, audio, essays, papers, and hands-on demos.
Featured podcastAudio
Give your AI a permanent mind
A 21-minute conversation on durable AI.
Listen now21:10
Flagship essay
Your AI Doesn't Need a Better Memory
It needs a cognitive layer: durable context plus attention that knows what should govern the task at hand.
Read essay →Continue the journey
A growing library of explainers, podcasts, papers, and demos.
Working paperPaper
Two Minds, One Room
What conversation becomes when both sides have a mind: the user gets a runtime, and disclosure becomes a task-scoped choice.
ArchitectureEssay
How a Mind Sleeps
The human mind is the muse: settling, replay, deep processing, REM-like discovery, waking, and proof that the work helped.
Working paperPaper
How Many Minds Do You Need?
A mind is a coherence boundary. Complex work wants focused minds whose attention has something coherent to organize around.
EssayEssay
Why we stopped saying "thinking"
One verb now covers two very different kinds of minds, and the conflation has costs. The case for a second word.
LexiconReference
Thinqing
People think. Digital minds thinq. The definition, etymology, and usage of the word for digital cognition.
ConceptEssay
The Interface Behind the Conversation
Where the interface is going: not another voice in the thread, but private attention over your own Mind.
Working paperPaper
Why Not Just Use a Folder of Markdown Files?
For one mind, markdown can be enough. Across identities, permissions, confidence, and task-sensitive attention, files stop being the whole answer.
Technical deep-divePaper
What Is a Mind?
Under the covers: each mind holds its own beliefs, anchored to a shared world of entities. The unit of memory is a belief held by someone, carrying a source, a confidence, and an audience. Why that is not a markdown vault plus embeddings.
Research paperPaper
Bigger Context Windows Are the Wrong Race
An 80-turn blind study on real production conversations. Proposed tight context plus retrieval matched full-transcript quality at about 15% of the uncached answer-prompt tokens - and brute-force context slightly increased hallucination.
Technical deep-divePaper
Agents Run in Their Own Context
Each agent runs isolated by default, with its own memory and model: a contained blast radius. How disclosure enforced at recall lets a private agent join a shared room without leaking.
Technical deep-divePaper
Bring Your Own Tools, Safely
How agents connect to real systems over MCP: self-bootstrapping OAuth, encrypted secrets, and a redacted per-call audit trail.
For developersGuide
A Shared Mind for Claude Code and Codex
Your coding agents forget everything between sessions. How thinqOS gives Claude Code and Codex one shared, persistent memory: recalled before each answer, captured after, the same across every tool and machine.
Technical deep-divePaper
Grows, But Under Authority
An agent's mind sharpens automatically; its identity changes only by owner-approved proposal. The honest answer to "but won't it go rogue?"
Technical deep-divePaper
One Mind, Many Compartments
Audience defines the compartment; who a memory is about is a separate index with no power over retrieval. Why that ordering is what makes an agent able to know you and unable to repeat it.
PodcastAudio
Securing the AI Cognitive Layer
A 19-minute conversation on compartmented agent memory: why audience, not subject, has to define the walls. Companion to One Mind, Many Compartments.
DemoDemo
See the Cognitive Layer in Action
A guided demo of thinqOS: context mobility, identity boundaries, and attention over a durable Mind.
ComparisonReference
What People Mean When They Say “Agent”
Four vendors, four incompatible definitions, and one thing missing from all of them. A dated, sourced comparison against the agent platforms — including where thinqOS loses.
Working paperPaper
What Is A2A?
The emerging standard for agents that have never met: what it specifies, what it leaves to the platform, and where thinqOS draws the line.
No published items match this filter yet.
Build a digital mind for yourself, your team, and your agents.
thinqOS is in private preview.