Field notes from the Age of Abundant CognitionBuild e803381 · 2026-09-25
Ryan HunterGet Ryan Hunter Daily

The Operator Deck · 24 cards

24 moves. Draw one.

Each card is an operator: a move you can apply to any situation, for the rest of your life. A fact helps you once. An operator keeps handing you new moves. Name the problem in front of you, draw a card, and ask its question.

Draw a card

24 in the deck

Shuffled. Your card lands here.

How to use the deck

  1. Write today's problem in one sentence.
  2. Shuffle. Draw one card.
  3. Ask its question about your problem. Out loud or on paper.
  4. Do the smallest move it points at before lunch.

The whole deck

Kernel · 01

Orient

Form a usable model of "where am I, and what's happening?"

Ask

Where am I, and what is actually happening here?

When
You feel lost, overwhelmed, or unsure which question even matters yet.
E.g.
Ask a model: "I don't know enough about this problem to know what questions matter. Help me map the space first."
The Operator Deck
Kernel · 02

Value

Distinguish what I care about and what I'd prefer.

Ask

What do I actually care about here, and what would I prefer?

When
You're busy on something and can't say why it matters to you.
E.g.
Two offers on the table. Before you compare the money, write down what you'd hate to lose. The ranking often flips.
The Operator Deck
Kernel · 03

Imagine

Represent states that don't exist yet.

Ask

What better states could exist that don't exist yet?

When
Something feels wrong, but you can only picture the current version.
E.g.
Ask a model: "Help me distinguish my current state, what feels wrong, and several possible better states without choosing for me."
The Operator Deck
Kernel · 04

Causalize

Ask what would have to become true for that.

Ask

What would have to become true for that?

When
You have a goal and a to-do list, and the list isn't moving the goal.
E.g.
"More revenue" needs more sales calls, which needs booked calls that actually show up. Stop at the first step you can do today.
The Operator Deck
Kernel · 05

Intervene

Turn one part of the causal frontier into an actual attempt.

Ask

Which one piece of this can I turn into a real attempt today?

When
The map is good enough and you're still mapping.
E.g.
You know booked calls don't show up. Today: send a reminder the morning of every call. One attempt, not a strategy.
The Operator Deck
Kernel · 06

Observe

Distinguish what happened from what I hoped would happen.

Ask

What happened, as opposed to what I hoped would happen?

When
You're about to call something a win or a failure.
E.g.
Ask: "What observable evidence would discriminate between these hypotheses?" Then go get that evidence, not more opinions.
The Operator Deck
Kernel · 07

Update

Change the model or the intervention rather than defend the story.

Ask

What does this change: the map, the move, or nothing?

When
The evidence disagrees with the story you've been telling.
E.g.
The reminder didn't lift the show rate. Don't defend it. Ask what else would have to be true, and change the move.
The Operator Deck
Kernel · 08

Acquire decoders

When blocked by ignorance: learn, ask, research, recruit, tool up.

Ask

Which decoder am I missing?

When
You're blocked by something you don't understand yet.
E.g.
Can't read a Japanese research paper? You don't need Japanese. You need a temporary bridge: translate it, then ask what prerequisite you're missing.
The Operator Deck
Kernel · 09

Persist

Keep enough state that an interruption doesn't reset the whole process.

Ask

If I stopped right now, what would I need written down to pick this back up?

When
Before you close the laptop, travel, or hand the work to someone else.
E.g.
End each session with three lines: what I wanted, what I tried, what happens next. Coming back costs minutes instead of a rebuild.
The Operator Deck
Kernel · 10

Recurse

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

Ask

What's blocking the process itself, and can I solve that the same way?

When
You're stuck, and "I don't know how" is the whole sentence.
E.g.
"I don't know how" becomes "what would let me learn how?", then "who or what can teach me?" That's where the rocket lights.
The Operator Deck
Operator · 01

80/20

Expect a few inputs to produce most of the output; find them and feed them.

Ask

Which few inputs are producing most of the result?

When
Everything on the list feels equally important.
E.g.
Sort last quarter's customers by revenue. Look at the top handful: what do they have in common? Go get more of that.
The Operator Deck
Operator · 02

The constraint

Find the one link governing throughput; effort anywhere else is waste.

Ask

What's the one thing limiting everything else right now?

When
You're working hard on many fronts and the total isn't moving.
E.g.
Twenty apparent problems become one underlying constraint. Not "we need more leads" but "the show rate is the bottleneck."
The Operator Deck
Operator · 03

The evaporating cloud

When two needs seem to conflict, find the assumption that makes them conflict, and test it.

Ask

This requires that only if what is true?

When
You're stuck between two things you both "have to" do.
E.g.
"To grow, I have to work more hours." Only if growth needs your hours. Break that assumption and a new move appears: hand off the doing.
The Operator Deck
Operator · 04

Expected value

Weigh outcomes by probability, not by vividness.

Ask

How likely is each outcome, and what is each one worth?

When
One vivid scenario, good or bad, is doing all your deciding.
E.g.
A one-in-ten shot at something worth twenty times the effort beats a sure thing worth one. The scary story isn't the odds.
The Operator Deck
Operator · 05

Feedback loops

Look for what reinforces itself and what corrects itself.

Ask

What here feeds itself, and what pushes back?

When
Something keeps getting better or worse on its own, and you don't know why.
E.g.
Better issues get more replies; replies tell you what to write next; better issues get more replies. Find the loop, then feed it.
The Operator Deck
Operator · 06

Intent is not authority

Conditional intent never authorizes an external action; approval covers exact text, recipient, channel, sender.

Ask

Did I approve this exact text, recipient, channel and sender?

When
Before you give an AI tool access to anything that talks to a person.
E.g.
"Ping them if needed" is not approval. The agent prepares the message; nothing leaves until you've read those exact words.
The Operator Deck
Operator · 07

Keep claims in their mode

A source saying X, someone believing X, a plan assuming X and X being true are four different things. Keep the label on.

Ask

Is this true, or did someone say, believe, assume or predict it?

When
An AI, or a confident person, hands you a conclusion.
E.g.
"The report says churn is about price." That's "a source says X". Check what the report was handed before you rebuild your pricing.
The Operator Deck
Operator · 08

Name the gain

Before you chase "more", name which kind you need: a route, a capability, a permission, resources or coordination.

Ask

Do I need a route, a skill, a yes, resources, or better coordination?

When
You feel stuck and you're about to buy a course or a tool.
E.g.
You don't need a new skill to launch. You need your business partner's yes. That's a permission, and no course will give it to you.
The Operator Deck
Operator · 09

No known path yet

Never conclude "impossible". Conclude "no known path under my current map", then search, ask, learn or test.

Ask

Is this impossible, or just no path I know yet?

When
You catch yourself saying "there's no way".
E.g.
No known way to reach that buyer? Four moves: search for how others did it, ask someone who has, get the missing fact, or run a small test.
The Operator Deck
Operator · 10

Opportunity cost

Price every choice by the best thing it displaces.

Ask

What's the best thing this choice pushes out?

When
A yes feels free.
E.g.
The extra client isn't just extra money. Its real price is whatever you'd have built with those ten hours a week.
The Operator Deck
Operator · 11

Reversibility

Move fast on two-way doors, slow on one-way doors.

Ask

If this turns out wrong, can I undo it?

When
You're stuck deliberating, or about to rush something.
E.g.
Changing a subject line: two-way door, just ship it. Signing a two-year lease: one-way door, slow down and check.
The Operator Deck
Operator · 12

Revise the lever

When nothing moved for two weeks, change a daily lever; don't change the goal, and don't change the system.

Ask

Which one daily lever do I change, keeping the goal and the system?

When
Nothing has moved for two weeks.
E.g.
Newsletter flat for two weeks? Don't change the goal or rebuild the system. Change one lever: the subject lines, the topic, or the send time.
The Operator Deck
Operator · 13

Standards, not dates

Turn a recurring task into a standard and a check; act only when the check fails.

Ask

What standard is this chore protecting, and how would I know it broke?

When
A recurring task keeps going "overdue" and making you feel behind.
E.g.
Not "OVERDUE: CLEAN COOP" every Saturday. Standard: the coop is clean enough. Check it weekly. Clean only if it's below the line.
The Operator Deck
Operator · 14

The weakest gate

Your ceiling is the weakest link in intent → … → learning, not your strongest capability.

Ask

Where does my loop leak: wanting, seeing, planning, acting, checking or learning?

When
You keep adding strength where you're already strong.
E.g.
10,000 agents and no way to check their work is industrial-scale hallucinated progress. The fix isn't more agents. It's observation.
The Operator Deck