Cognition is becoming abundant. Agency isn't.

Here's you. Now. Physicists draw a moment like this: a point, the past trailing below it, and above it a cone of every future that can still be reached from here.

Most people think there's one cone. There are three.

The actual cone

Every state truly reachable from the current world under the actual laws, resources, actors and institutions. Nobody knows it perfectly; strategy always runs against a model of it.

Every dot is a future. A different job, a finished book, a company, a Tuesday that goes differently.

The epistemic cone

Inside it: the futures you believe are reachable. Your map. Everything you'd ever plan for lives in here, because you can't aim at a future you can't picture.

The effective cone

And inside that, the small bright one: what you can actually cause, with the skills, permissions, money, people and hours you have right now.

That's your agency. Not what you know. Not what you could imagine. What you can move.

The future you can't see

Say reality has a path from here to Z. Somebody knows a trick you don't. Z is real; it's in the big cone. It just isn't on your map.

So "that's impossible" is almost never true. The honest sentence is: "I don't know a path to that yet." The first one ends the conversation. The second one is a to-do: search, ask someone, find out, run a test.

What AI actually did

Cheap machine intelligence blew your map wide open. Ask a model anything and you can suddenly see routes you never knew existed. Z is on the map now.

Your bright cone grew too, but less, and only where you acted on what you saw. The difference is the gap. It's why you can feel more capable than ever and still be stuck on the same things as last year.

Five ways to widen the bright cone

"I can do more now" hides five different things. Name which one you're missing:

1. A route. Someone shows you a path you couldn't see. (widens your map)
2. A capability. You learn to do the thing, or get a tool that does it.
3. A permission. You're allowed to act: the role, the sign-off, the license.
4. Resources. The money, time or materials the path costs.
5. Coordination. The other people moving with you.

The first widens what you can see. The other four widen what you can cause. They are not the same fix.

Somebody spends twenty years figuring something out. The failures, the conversations, the experiments, the nights it didn't work.

All of that lives in one head. When that head stops, it's gone.

Writing converts it

Then they write it down. Writing is a converter: it turns thinking, which evaporates, into a thing that lasts.

The book is a carrier. It holds the potential to change what someone can do. Not the author's life. A compressed shape of what their life taught them.

And then it waits

A carrier can sit for ten years and do nothing. The agency is in there, frozen, waiting for someone who can unlock it.

Publishing copies it

Publishing converts one intervention into a million. Same words, a million shelves. Publish once and distribute to millions.

Reading unlocks it

Then you read it, and a route that was invisible to you becomes visible. Nothing in the world moved. Your map did.

Deciding and doing spend it

You decide. You act. The world is different. Twenty years of someone else's search, spent by you on a Tuesday.

That's the whole trick of civilization: somebody pays for the discovery once, and everyone after gets it cheap.

Every arrow here is a transducer: something that turns agency from one form into another. Writing, publishing, reading, a team, a contract, money, code, an AI model.

Here's everything a great book could do for you, pouring down from the top. Between the page and a changed life there are eight gates. The light has to get through all of them.

1. Access. It reaches you.
2. Recognition. You know it's for you.
3. Decoding. You can read its language.
4. Comprehension. You rebuild the meaning.
5. Applicability. You see where it fits your life.
6. Capability. You can do what it describes.
7. Authority. You're allowed to.
8. Execution. You actually do it.

Hand it to an alien

Give a visitor from another planet every book humans ever wrote, on one drive. They don't know what the drive is, or electricity, or English, or books.

One gate is shut. It doesn't matter how open the others are. The agency they received: about zero.

Now give them a translator

A machine that turns the drive into their own language and their own ideas. Suddenly the light pours through.

Nothing on the drive changed. The coupling changed: how well the reader could connect to what was already there. That's what AI models are for most people. The cheapest translator in history.

Now you

You read the book. You understood it. You even agreed with it. Top four gates, wide open.

Then: could you actually do it? Were you allowed? Did you? The gates multiply. A strong page times a shut door is a shut door. Your weakest gate is your ceiling, not your strongest one.

Books, models, mentors: all carriers. What turns them into a changed life is a loop you run. Here it is, seven moves:

1. Orient. Form a usable model of "where am I, and what's happening?"
2. Value. Distinguish what I care about and what I'd prefer.
3. Imagine. Represent states that don't exist yet.
4. Causalize. Ask what would have to become true for that.
5. Intervene. Turn one part of the causal frontier into an actual attempt.
6. Observe. Distinguish what happened from what I hoped would happen.
7. Update. Change the model or the intervention rather than defend the story.

Then you hit a wall

You go to intervene and you don't know how. The paper is in Japanese. The tool has no manual. The person who knows won't answer.

Most people stop here. So do a lot of AI agents: they hit something they don't know and either make something up or quit. Then someone else has to come restart the loop.

Unless you've got two more moves

8. Acquire decoders. When blocked by ignorance: learn, ask, research, recruit, tool up.
9. Persist. Keep enough state that an interruption doesn't reset the whole process.

"I can't read Japanese" becomes: translation exists → a model can translate → the physics is still confusing → ask for what to learn first → learn it → come back.

You never needed Japanese. You needed a way to borrow a bridge across it.

Recurse

Apply the same process to the limitations of the process itself.

"I don't know how" becomes "what would let me learn how?" And when that stalls: "who or what could teach me to learn it?"

That's the threshold. Below it, someone else has to keep restarting your loop. Above it, almost every block turns into another problem you can solve.

You can run these today

They're not magic prompts. They're handles on the loop. Paste one into any model when you're stuck:

• OrientI don't know enough about this problem to know what questions matter. Help me map the space first.
• ImagineHelp me distinguish my current state, what feels wrong, and several possible better states without choosing for me.
• CausalizeRepresent the outcome causally and stop decomposition when we reach something I can actually do.
• ObserveWhat observable evidence would discriminate between these hypotheses?
• Acquire decodersTell me what prerequisite decoder I'm missing.

Remember the twenty years inside the book? Civilization is a machine for not paying that twice. Someone pays for the discovery once; everyone after inherits it.

Every big technology made inheriting cheaper. Books stored it. Search found it. Phones put it in your pocket.

Language models cut the cost of decoding: the right idea, in your words, at your level. Agents are starting to cut the cost of acting on it.

So the headline isn't "AI is powerful." Everybody's heard that. It's this: you just got the cheapest translator ever built for everything humanity figured out, and most people haven't learned to plug themselves into it yet.

The row nobody's built

Nobody has cut the cost of keeping it together: holding the goal, remembering what worked, keeping five tools and three people and a month of effort pointed the same way.

That row is still yours. Which means it's where most of your agency leaks, and where the biggest edge is.