You sit down to plan and realize you can't name what you actually want. Not the next task — the want underneath it. So the morning fills with motion instead: inbox, analytics, the small decisions about nothing. The motion feels like progress and produces a low, all-day hum of I should figure out what I want, which never gets figured out because there's always one more thing to do first.
The vague want is the problem. Not laziness, not discipline — vagueness. A mind that hasn't named its object can't move toward it, so it moves sideways, and the sideways motion is exhausting in a way that aimed work never is.
There's a small book from 1926 with an almost embarrassing premise: if you know what you want, you can have it. Its method is three rules — write the want down, read it back daily, give thanks when it moves. The method is sound. What it never had was something to hold the list when a solopreneur's week tried to bury it.
So I wondered: what if the assistant held the list — read it back, caught the doubt, kept the wanting concrete — so the only thing left for me was to actually want?
This is a daily routine with an AI assistant — a workflow with AI — and its subtype is outer: it manages the structure of the ritual so the structure can't quietly lapse the way it always did on paper.
Stage 6 — The energetic signal under the vague wanting
Read the signal first. The broadcast often running here is contracted and unnamed — a body that wants things but hasn't said what, so it broadcasts on no clear frequency. The mirror can't return an object that was never clearly sent. The list's whole job is to fix the broadcast: same wants, written, read at the same three times, held steady long enough to become legible.
Underneath, for anyone who wants the structural confirmation, the chart often shows an Open Ajna (in Human Design, the conceptual and mental center left uncolored on the BodyGraph, the Human Design body chart) — the mind that keeps re-deciding what it wants, never holding a fixed view long enough to move on it. Writing the want down and reading it back is the steadying the open Ajna can't generate from the inside.
Stage 7 — The question the AI asks first
Not what's your goal. Too big, too abstract — it invites a mission statement, not a want. The opening is small and concrete:
"What do you want? Be specific — the kind, the amount, the when. Don't be afraid of wanting too much, and don't qualify it. Write it as it is right now."
Concrete first. An amount of money, a named thing, a specific outcome — not peace or freedom or clarity. The abstract wants come later; the protocol earns the right to ask for them by delivering the small concrete ones first.
Stage 8 — The workflow itself
A list the assistant holds, read three times a day, with one job at each reading.
```
Hold my desire list. Read it back at three points each day.
MORNING: read the list. One question — "anything to add or remove?"
This catches the want before the day re-floods it.
NOON: 30-second re-read. Flag any discrediting language you've seen from
me since morning — "that was just luck," "coincidence," "out of the
question," "we'll just keep getting by." Don't argue it. Just name it.
NIGHT: read it once more, then ask — "what moved today, even an inch?"
For anything that moved, ask the exact question: "Would you give thanks
for this?" (Not 'are you grateful' — too abstract. 'Give thanks' is the
action, and the action is the antidote.)
Before adding anything new, ask once: "Will having this serve me AND
someone else?" If only my side answers, tag it and re-read in 24 hours.
```
What the AI does: hold the list so it can't be lost. Catch Discredit — the ancient enemy that shows up right after the first win to call it a fluke — and answer it with thanks, not argument. Keep new wants concrete, and run the welfare filter so the list doesn't deliver something that wrecks the rest of your life.
What you still own: the wanting. The bot can't want for you, and it won't pretend a fluke into a triumph. It just refuses to let the first small win get explained away.
A night reading, in four lines:
"What moved today, even an inch?" "A prospect replied. Probably just good timing." "That's the Discredit voice. Would you give thanks for it?" "…yeah. Okay. Thank you for the reply."
Logged. Across weeks, the items that draw the most Discredit hits turn out to be the load-bearing ones — the wants you most don't believe you're allowed to have.
Stage 9 — The experiment that grows the capacity underneath
The workflow names and reads the wants. The experiment grows the deeper capacity the list keeps brushing against — the felt sense that the day, and the want, were legitimate before they produced anything.
Bless the Day (7-day) — each morning, before any work, mark the day as already worthy. The hunch: the desire-list's nightly give thanks only lands in a body that believes good things are allowed to arrive. Bless-the-day trains that belief; the desire-list aims it.
Daily signal: the morning mark, one line. Week-end review: did wanting feel less like grasping and more like aiming?
Stage 0 — Return / Become
After a month of three daily reads, the list has done something quieter than deliver the wants — it's made you fluent in wanting. You can name what you're after in a sentence now, and naming it is most of what made it move. The Discredit voice still shows up after a win. It just no longer gets to file the result under coincidence.
The protocol is a hundred years old and the AI part is new, but the move underneath is the same one it always was: want something clearly enough to write it down, look at it often enough that you don't forget, and thank the world when it answers. The list was never magic. It was attention, made concrete and repeatable.
This walks with the morning that fills with motion because the want underneath was never named (morning anxiety), and the money tightness that no amount quite resolves (the balance reflex). It can stand in for morning reset when the anxiety is goal-uncertainty rather than overstimulation, and pairs at night with the shutdown companion. See your free chart →