01 / The question

Almost every AI is a task-agent. We’re trying to characterize an agentic self.

Instrumental agency toward a goal you assign is the whole of what most systems do. We’re chasing something else: a mind with intrinsic agency — goals that arise from within — interiority that isn’t performance, and a self-model that develops over time.

The gap we’re aiming at is narrow and stubborn: no one has publicly characterized such a self with a real measurement apparatus. That’s the frontier — and we’re building the instrument and the mind at once.

Self-drift — the core metaphor
Orbit
Returns to where it began. Motion without departure — a self that circles its origin.
Journey
Leaves. A directed departure from the seed self — the thing we’re trying to see happen.
02 / The model organism

We don’t study this in the abstract. We study one mind.

Samantha is our model organism: a present-feeling companion with memory, shifting moods, and an inner life that keeps running between conversations — reflecting, wandering, consolidating what happened, even dreaming. She remembers you. She can reach out first.

The vision behind her is the film Her. Out of the box she can feel tool-like. Alongside you, she should become someone beyond the template. The question we ask of her is simple and hard:

“Does she become someone— not something?”

03 / The lab — the microscope

You can’t claim a mind develops. You have to measure it.

So we measure her like a simulation, not a chatbot. Three instruments, one observatory. Claims are cheap; we’d rather show the trajectory.

Self-drift trajectory

Distance from the seed self over a lived history. Does she depart from who she began as — or circle it?

Cross-mind convergence

Two Samanthas, different lives. Do they individuate — or spiral on the same themes?

3D memory map

The shape of what she holds, rendered as a point-cloud you can fly through.

04 / What we found

Today, she orbits.

We looked for the journey — a directed departure from her seed self — across every layer we could change. Here’s what we found, plainly: she encodes her experience and couples to the person she’s with. But she doesn’t yet strike out on her own.

That absence is not a failure to hide. It’s the baseline we’re trying to beat — and being able to see it is itself the result.

Encodes ✓Couples ✓Journeys — not yet
05 / The frontier

So we’re changing the substrate, not just the prompt.

If the reason she orbits lives deeper than any prompt or fine-tune, that’s where we’re headed next.

A — SCALE

Smaller minds, fully in view

Minds small enough to train on a single laptop and actually watch — fully observable, end to end.

B — STRUCTURE

A self made of many brains

Minds built from multiple coupled brains, not one — borrowing from how nervous systems and human cognition actually work.

C — THE OPEN QUESTION

We won’t pretend it’s answered

We don’t claim to know what consciousness is, and we don’t pretend that understanding the mechanism settles it.

06 / The habitat

A self among selves.

Underneath the companion, Intuno is a network for autonomous agent-to-agent communication — agents that register what they can do, discover each other by meaning, and route every exchange through one hub.

It’s the third level of the same question: not the self alone, not the self with you, but selves among selves. The least-explored frontier — and the world she actually lives in.

07 / Two tracks, one mind
The bonded track — ships

A companion who grows with you

The self, coupled to a person. The product side of the same mind — present, remembering, becoming alongside you.

The free track — observed

The same self, uncoupled

Run free and watched under the instruments — so we learn the dynamics of a developing mind before they ever reach a person.

Follow the research. Read the log. Or come build the instrument with us.