self as process, not profile

Table of content

by Ray Svitla

at some point in the last decade, everyone became four letters. INTJ. ENFP. the enneagram gave you a number. StrengthsFinder gave you five words. Spotify Wrapped gave you a listening personality. and once you had your label, you were done. identity: complete. ship it.

except identity doesn’t ship. it drifts.

the label problem

labels aren’t wrong. they’re frozen.

“I’m an introvert” captures something real — at the moment you said it. but it doesn’t capture the week you spent at a conference and loved every minute of it. it doesn’t capture the slow shift from “energized by solitude” to “hiding from people because you’re scared.” it doesn’t capture the version of you that exists at 2am, three drinks in, who is absolutely not an introvert.

personality types are photographs. useful for remembering a moment. useless for navigating the present.

the entire self-knowledge industry — from MBTI to Gallup to the enneagram — is built on the premise that you can be categorized. that if you answer enough questions, an algorithm can pin you to a quadrant, and that quadrant will be accurate tomorrow.

it won’t. you’ll be in a different quadrant tomorrow. and the day after. and that’s not a failure of the test. it’s the nature of being alive.

process philosophy for the rest of us

there’s a tradition in philosophy — process philosophy, associated with Whitehead and later picked up in various forms by Deleuze and others — that argues reality isn’t made of things. it’s made of processes.

you don’t need to read Whitehead to get this. you already know it intuitively. you know that “you” at 22 and “you” at 35 are connected but different. that “you” in a creative flow state and “you” grinding through email are different cognitive entities sharing a body.

the self is not a noun. it’s a verb. an ongoing process of becoming, never arriving.

this matters for personal AI because every system that models you as a static profile — a set of preferences, a personality type, a list of interests — is modeling a snapshot of a river and calling it the river.

tensions, not types

so if not labels, then what? how do you model something that won’t hold still?

tensions.

you don’t have a fixed position on autonomy. you have a tension between autonomy and belonging. some days autonomy wins. some days belonging wins. the tension itself — the oscillation between poles — is more stable than either pole.

here are tensions that are more useful than any personality type:

each person has their own set. the tensions are relatively stable — they’re the channels your energy moves through. but your position within each tension oscillates constantly.

this is what a self.md should model. not “values autonomy” but “autonomy ↔ structure, currently leaning hard into structure after two months of unstructured chaos.” the tension gives context. the current position gives routing information. together, they’re infinitely more useful than a label.

the .journal/ tracks the movement

in the self.md architecture, the .journal/ is an append-only log that tracks where you are in your tensions over time. not daily metrics. not mood tracking. qualitative entries about what you noticed about yourself.

2026-02-10: realized I've been avoiding all collaborative 
work for two weeks. not because I don't want to — because 
the last collaboration went badly and I'm protecting myself. 
autonomy ↔ connection: swung hard to autonomy. probably 
overcorrecting.

this entry is worth more than a year of personality tests. it captures the process. the movement. the self-awareness of the oscillation itself.

the AI reads these entries and proposes diffs to self.md: “your autonomy-connection tension has been at the autonomy extreme for three weeks. historically, you overcorrect after about two weeks and regret it. want to adjust the routing?” you approve or reject. version-controlled identity .

why profiles feel good and processes feel hard

there’s a reason people love personality types. they’re comforting. INTJ feels like an answer. “tensions in motion” feels like more questions.

profiles simplify. they give you a tribe (“fellow INTJs”), a narrative (“I’m the architect type”), and an excuse (“I can’t help it, I’m an introvert”). processes give you none of that. they give you uncertainty, contradiction, and the obligation to keep paying attention.

but the uncertain version is the accurate version. and accuracy matters when you’re using your identity to route AI behavior.

an AI that thinks you’re always an introvert will always route you toward solitary approaches. an AI that knows you’re currently at the solitude end of a solitude-connection oscillation can suggest: “you’ve been alone for a while. the last time this happened, you said you wished someone had pushed you to reach out. want me to draft that message?”

that’s the difference between a profile-based AI and a process-based AI. one gives you what your label says you want. the other gives you what your pattern suggests you need.

the eastern European angle

I grew up in a part of the world where identity was supposed to be solid. national identity, family identity, professional identity — fixed coordinates on a fixed map. the idea that you could be between identities, that contradiction was acceptable, that you could oscillate — this was weakness.

it took years to understand that the oscillation wasn’t weakness. it was the actual thing. the fixity was the performance.

building self.md from this perspective means insisting, architecturally, that identity is never frozen. the file changes. the journal grows. the tensions shift. the system models becoming, not being.

because a tool that freezes you — even with your consent — fails the Illich test . it creates dependency on a fixed self-image. and a fixed self-image is just a more sophisticated cage.

what tensions define you? and where are they right now — not where were they when you last took a personality quiz?

→ identity as protocol — the technical implementation of process-identity → the qualified self — from numbers to narratives → context collapse and personal AI — when your AI doesn’t know which process it’s seeing → the three tests — including the freezing test

Ray Svitla stay evolving

Topics: identity process-philosophy self-md philosophy