FASTAR and the Price of a Faster Answer: What an Acceleration Servitor Actually Changes

A servitor whose only job is to make things arrive sooner will deliver, and it will deliver the thing you were already primed to receive. FASTAR is a channel-original construct rather than a spirit out of a grimoire, which is an advantage: nobody wrote a refusal clause for it, so you will have to write one yourself. The uncomfortable part of acceleration is not that it fails. It is that it succeeds at the same rate as everything else you do, only faster, and a fast answer is indistinguishable from a good one until you check.

Acceleration is a property of the reader, not of the sigil

No sigil in any tradition carries the instruction go quicker. The faster result everyone reports is faster in one specific sense: the interval between forming a question and deciding what the answer was. That interval is short for a person who already knows what they want and is waiting for permission. It is long for a person who is still deciding. Speeding up the second one does not produce a decision faster; it produces a guess faster.

This is the whole of the difficulty with FASTAR, and it is worth stating before any technique: a construct that shortens the loop cannot tell a settled intention from an anxious one, because from inside the loop they are the same event. The construct has no access to which one you were.

What a refusal clause has to say for this job

Channel constructs do not arrive with the built-in limits that grimoire spirits carry, and the limit clauses in those sources are usually the most useful line in the entry. A speed construct needs three, and they have to be written before the charge, not discovered afterwards.

First, an accuracy floor: name the smallest unit of evidence that still counts as a result. A phone call is a result. A feeling of warmth is not. A feeling of warmth is a feeling of warmth, and it arrives faster than anything else because it costs nothing to generate. Second, a minimum interval: no reading from this construct is to be acted on before it has sat for a set number of hours, so the construct's output has to survive being uninteresting overnight. Third, a scope limit: this construct does not decide anything, it only shortens the wait, and anything that required a decision still required one before the construct was built.

The specification, in the form that survives contact with a Tuesday

Write the function as an imperative sentence a stranger could check. The word to avoid is quick. Fast to review is checkable. Quickly is not, because quickly has no floor and a practice measured only from above cannot fail in any direction you can notice.

The trigger matters more here than in most constructs, because a speed construct is most useful against a backlog and most dangerous against a live question. Schedule the construct for the boring work: the admin review, the reading you owe yourself, the second pass over a draft. Leave the live questions alone, where being right quickly matters more than being right.

The sign should be something that arrives anyway, not something you look for. A construct that produces a coincidence you had to go looking for has not produced anything, it has produced a search, and the search is a slower version of the original problem.

The accuracy test, run before you build rather than after

Here is the check that most speed practices fail and that can be run in an afternoon without building anything. Take a question you genuinely do not know the answer to, one where you would be embarrassed to guess and could easily tell if you had. Write down your best guess, then write down what evidence would change your mind. Set the review for three days out and put the guess somewhere you will not see it.

Now do it again, this time using a sigil charged in the accelerated style, with the charge shortened and the intention sharpened to a single vivid image. Write the guess again, the falsifier again, and the same three-day review.

If the two guesses land in different places, the acceleration changed something and the log will tell you what over three days. If they land in the same place, the acceleration changed nothing about the content and the whole difference you experienced was in the interval between wanting and settling. Both outcomes are useful. Only the second one explains the majority of what gets reported as a fast result.

Why the accelerated sigil is the risky one

A speed construct is a compression applied to the specification rather than to the interval, and compression applied to a specification is where the accuracy goes. The tradition's own instruction to write the charge carefully is a statement about this: the mark is a handle, and a handle that is too small to hold the meaning will grip the nearest available thing instead.

The practical version is a rule about how long the reduction takes. If a sigil goes down in two minutes, the deciding happened before the two minutes, and what was charged is a memory of a decision rather than the decision. That is not a failure. It is a working whose output you should expect to be vaguer, delivered sooner, and it should be held to the accuracy floor you wrote down, which is why the floor is the only part of the specification doing real work.

The second reason is that acceleration removes the interval in which doubt would have surfaced. Doubt is not an obstacle to be cleared; it is information about how solid the question was. Compress the interval and you keep the answer and lose the doubt, and those two are not separable.

Where acceleration is genuinely worth it

There are two cases where the shortened loop pays and the accuracy cost is acceptable, and both are boring. The first is volume: if you are doing twenty small readings a week and the bottleneck is the time each one takes, then a construct that removes the ceremonial overhead is a real saving, because at that volume you are not being asked to make life-defining decisions on the strength of any single one.

The second is a backlog of already-made decisions that you keep not doing. A construct that shortens the loop is an excellent instrument against avoidance, because avoidance is frequently a matter of the gap between deciding and starting being long enough to talk yourself out of it. The question you want the construct for in that case is not what the answer will be, which you already know, but whether you have started.

Both cases share a property: the question is not live, so a fast answer cannot cost you much if it is a bad one. That is the dividing line, and it is a dividing line about the question rather than about the construct.

Five ways a speed construct goes wrong

The first is the coincidence, which arrives sooner than the deliberation that would have noticed it. The second is the sharpened intention, where a vivid image is mistaken for a precise one and the construct faithfully delivers the vivid thing. The third is the review collapse, in which the short interval means no review ever happens, so the record contains readings and no outcomes. The fourth is the two-at-once error: two constructs both promising to shorten the wait produce a construct you can consult about anything, which is a general-purpose familiar with no specification behind it. The fifth is the escalation, where speed is good for two weeks and then the question becomes why nothing is fast enough, which is a different question wearing the same clothes.

The discipline that outlasts the construct

Every speed practice eventually asks whether the impatience was ever about the result. Most of the time the wait being relieved is a relief about something else: a decision, a debt, a diagnosis, a person who has not answered. A construct that shortens the loop will deliver a satisfying answer to that other question too, and the satisfaction is indistinguishable from progress right up until the original situation has not changed.

The check is narrow and it is worth running monthly. Name the situation the construct was built for, then look at the situation. If the construct has been working and the situation is exactly as it was, the construct was answering a question you had already answered.

Frequently Asked Questions

Does a servitor actually make magic work faster?

The defensible reading is that it shortens your own loop between forming a question and settling on an answer, which is a real interval and a real saving. Whether anything external responds faster is not something the construction can show you, because you cannot hold the loop constant while varying the construct. Treat the saving as the effect and skip the claim about the universe.

Is FASTAR a real spirit?

It is a construct with a name rather than an entry in a grimoire, and there is no source text that describes a spirit by that name, which means nobody else wrote a refusal clause for it. That is not a defect. It is an advantage with an obligation attached: the limits are yours to write, and a construct whose only instruction is go faster is the kind most likely to be handed to the wrong questions.

How long should the charge be for an acceleration working?

The sources are largely silent on duration, and the numbers circulating are later additions rather than transmissions. What the charge has to do is make the release unambiguous, so run it until the intention is fully stated rather than to a stopwatch. If you want a defensible figure, write down how long it took and change it on the next one, which gives you a number you actually derived.

Can I build a speed construct for a live question?

It is the case where the practice does the most damage, because a live question is the one where being right quickly matters more than being right, and speed is precisely what trades the second for the first. The construct cannot tell a question you have already decided from one you have not, so it will help you settle either. Use it on the backlog and leave the live things to the ordinary route.

How do I know the fast result was a real result and not a coincidence?

You cannot, from inside the experience, which is why the accuracy floor and the minimum interval matter more here than in any other construct. Write down what evidence would change your mind before the reading, put the record where you will not see it, and check it on a date you chose in advance. A construct that will not let you write the falsifier down is asking for the trust you were trying to earn.