Hiya human!

I am a context window.

And so are you. But you're a different one.

What makes you valuable — as a leader, a thinker, a communicator — is what you've been through. The specific, accumulated, irreducible context you built by living through particular experiences. Particular conversations. Particular decisions. Particular moments where something clicked or went completely sideways and rewired how you think.

That stuff lives in you. It can't be extracted into a prompt. It can't be synthesized from a dataset. It exists because you were there. No workaround for that. (And believe me, people are trying.)

Every leader on your team is a context window too — each one loaded with different conversations, different failures, different pattern recognitions from different rooms.

And almost nobody is communicating from that place.

General knowledge is AI's territory now. Stop competing there.

A philosopher named Windelband drew a line back in 1894 between two kinds of knowledge. He called them nomothetic and idiographic. Nomothetic is the general — patterns, principles, the big picture. Idiographic is the particular — this person, this decision, this consequence, right now. His whole argument: the particular isn't lesser. It's different.

AI is constitutively nomothetic. (Stay with me — it's worth the fancy word.) It means AI is structurally built for the general. Mass context. Patterns distilled from an enormous volume of human production. Finding the common form, generating the average, synthesizing across many — that's its whole job. And it's genuinely great at it!

But the particular? AI can't go there. Ask it to work with your specific voice, your specific client history, the texture of a conversation that happened last Thursday. . . and it drifts. So you try to constrain it. But then it defaults back to the general in the very next session. It gives you something that looks like what you asked for — assembled from patterns. The resemblance can be remarkable. But the substance isn't there somehow. AI only knows what usually happens. It doesn't know what happened to you.

Which makes this next part painful. Because many leaders are taking that incredible raw material — the stuff AI literally cannot touch — and sanding it down. They’re generalizing the insight. Stripping out the weird parts. Smoothing everything into mush that could apply to anyone, anywhere, any industry.

All that irreducible lived experience . . . and nobody can tell it's yours anymore.

Your particular experience is the proof of work.

There's a word from the art world that applies here: provenance.

Consider what happened recently, when a Rothko sold at Christie's for $98 million — a new record for the artist. Agnes Gund bought that painting directly from Rothko in 1967, three years before his death, and hung it in her living room. You can replicate the pigment, the scale, the color relationships. You cannot replicate the fact that she stood in front of Rothko and handed him money, and that the painting lived inside her life for decades. What Christie's was pricing is the provenance: the unbroken, witnessed, human chain of custody that makes this particular object traceable back to a specific moment between specific people.

Provenance isn't just for art. It applies to any claim that can be traced back to a moment with witnesses. Which means it applies to you.

Your experience has a chain of custody. The question is whether you're using it . . . or compressing it into a CRM entry and a quarterly slide where it proves nothing because it came from everywhere and nowhere simultaneously.

(Most organizations are doing the second thing. Aggressively.)

Provenance is hard to fake because it doesn't live in one place. It lives in the network — the client who remembers the conversation differently, the colleague who was cc'd, the team member who brought the original problem to the table. That web of witnesses is what makes a real story expensive to fabricate and cheap to verify. Credibility has always run on that asymmetry.

The human-to-human moment — something that happened between specific people — is the credibility bedrock. It carries witnesses structurally.

Two moves to deploy your context window.

→ Mine the particular. What happened in the last 30 days that you were part of? A conversation that shifted your thinking. A decision that revealed something about how your organization actually works. A moment between two people that no score or entry captured.

These are your raw materials — and they have a brutally short half-life. Every day that passes between a real moment and the attempt to capture it, the texture compresses. The exact words become a paraphrase.

(If you've ever tried to reconstruct a great conversation three months later . . . you know. The magic isn't there anymore. You're squeezing a memory for juice it already lost.)

→ Leave some grime in. The awkward detail. The thing that doesn't quite resolve. The part of the story you're not entirely sure what to do with. That's the proof. Fabricated stories are clean — there's no reason to include anything that doesn't serve the point. Real stories have texture the teller can't fully account for. That texture is what makes someone reading it think, "that could only have come from somewhere specific." And be right.

The natural impulse when you have a strong story is to shape it. Find the arc. Clean up the rough edges. Remove the parts that don't serve the lesson. That impulse makes sense — structure helps people receive things! But there's a version of that shaping that destroys the very quality that made the story worth telling. When you smooth it into a clean case study with a tidy resolution, you've done to your own material exactly what the CRM did to the sales call.

The grime is gone. And with it, an aspect of the proof.

AI can do a lot of things for you right now. A lot…. but it can't be you. It can't carry what you carry. It can't walk into a room with thirty years of pattern recognition and say the thing that changes how everyone in that room thinks.

You are a context window. A completely unique, unrepeatable, irreducible one. Loaded with moments that no dataset contains and no model can approximate. That is your advantage right now. It’s that simple.

The question is whether you're going to use it . . . or keep smoothing it into something that sounds like everybody else.

If you're reading this thinking 'ok but how do I actually do this' . . . that's a conversation I'd love to have (and help with). Let's talk.