AI Horary Astrology: What Software Can—and Cannot—Judge
Horary software can calculate a chart, apply explicit traditional rules, and help explain the evidence. Here is where deterministic computation ends and AI begins.
In questa guida
Horary astrology was codified in the 1600s by William Lilly. Many of its operations are explicit: calculate the figure, assign significators, measure aspects, examine dignity and reception, then weigh perfection and denial. That makes horary unusually suited to software — but “software” and “AI” are not the same thing.
The short answer
Software can calculate a chart and show the rules it applied. AI can help explain that information in ordinary language. Neither one knows hidden facts, predicts the future with certainty, or replaces your judgment. Think of the calculator as the measuring tape and AI as the person helping label the measurements — neither is the building itself.
| Software can do | It cannot do |
|---|---|
| Calculate astronomical positions and chart relationships | Know someone’s private thoughts or future choices |
| Apply declared, inspectable rules consistently | Turn symbolism into a guaranteed outcome |
| Explain structured chart evidence more accessibly | Replace medical, legal, financial, or safety advice |
The astronomical chart can be calculated deterministically. Traditional rules can be evaluated in code. AI can help classify a complicated question or turn structured evidence into readable prose. Those layers have different strengths and different failure modes.
What horary software actually does
A horary reading breaks into two very different jobs:
- Calculation: given a moment and a place, where are the planets? What houses do they fall in? What aspects do they make? Which planet rules which sign, and what are its dignities and debilities?
- Judgment: given that chart and a clearly scoped question, what testimony supports or denies the matter?
The first job is pure arithmetic. Swiss Ephemeris and its open-source siblings have been solving it flawlessly for decades. Any credible horary tool — AI or otherwise — is using the same underlying astronomy that professional astrologers use. There's no room for mysticism in planetary positions; the Moon is where the Moon is.
The second job is where it gets interesting. Traditional horary judgment follows a procedure:
- Identify the significators — which planets represent the querent, the thing asked about, and any third parties
- Check the Moon as co-significator and as the describer of events
- Look for perfection — a clean aspect between significators that brings them together
- Check for denials — aspects that block, combustion, retrograde turns, sign changes before perfection completes
- Weigh reception — does the other party regard the querent, or is the interest one-sided
- Consider timing — when does the perfecting aspect become exact, in what unit of time
- Apply considerations before judgment — radicality checks that determine whether the chart is safe to judge at all
Many of these operations are structured and computable. They do not need a language model. The difficult parts are deciding what the question truly refers to, resolving edge cases where authorities differ, and writing a proportionate judgment from the evidence.
What computation does well
Three things, honestly:
Speed and accuracy on the mechanical work. A human astrologer computing a horary chart by hand takes minutes to hours. A well-built horary engine identifies every aspect, every reception, every dignity, and every mutual reception in under a second — and it doesn't miss the fiddly ones. Mutual reception is easy to overlook when you're tired. Software doesn't get tired.
Consistency across thousands of charts. Every human judges slightly differently. A deterministic rules engine can apply one declared standard the same way each time. That consistency is useful for learning because you can inspect which rule fired and compare it with the source tradition.
Surfacing every testimony. A traditional judgment weighs testimonies — pieces of evidence that support, deny, delay, or qualify the matter. Software can expose the strong, the weak, and the contradictory ones together, so the human reader can see the case rather than only a conclusion.
What AI doesn't do well
Just as honest: there are parts of horary where AI is weaker than a trained astrologer.
Reading the querent. Horary is famously unforgiving of dishonest questions. A good astrologer asks you three questions before you ask your one. What's really going on? What do you already know? What are you afraid to hear? A computer can't do that — it judges the question as asked, not the question beneath the question.
Context that isn't in the chart. Charts don't know you. If you ask "will I get the job?" and forgot to mention you haven't actually applied yet, the chart will describe whatever situation is active in your life — which may not be the job you had in mind. A human astrologer would catch that in conversation. Software judges what's there.
Ambiguous aspects and minor dignities. Is a separating aspect by 30 minutes still counted? Is a planet at the last degree of a sign still "in" that sign for judgment purposes? Horary has edge cases where tradition divides. Lilly says one thing, Bonatti says another, Masha'allah says a third. Software has to pick one rule and stick with it. That's a feature (consistency) and a limitation (inflexibility) at the same time.
Knowing when to refuse. A great horary astrologer will sometimes decline a question: you've asked this three times, the answer isn't changing. This isn't a horary question, it's a therapy question. Come back when you're actually at the crossroads. A computer will answer every time, even when the right move is silence.
How to use an AI-assisted horary tool well
The most productive use is to treat the calculated chart and explicit testimonies as the inspectable foundation, and any AI-written reading as a drafted explanation of that evidence, not an oracle beyond it.
Ask real questions. Horary is not a search engine. Don't ask what you can Google. Ask what's pressing on you that you genuinely can't resolve on your own. Will this offer come through? Is this person telling the truth? Where is the thing I lost? Specific, time-bound, unknown.
Ask once. One of the oldest horary rules: don't ask the same question twice. The chart answers the moment you ask. Re-asking produces a chart of your impatience, not your situation. If AI makes asking cheap, the temptation to spam is real — resist it.
Read the evidence, not only the prose. Pay attention to which significators are involved, what they are doing, what the Moon is doing, and whether the stated conclusion follows. That is how you learn to read charts and how you catch a weak explanation.
Do not use timing as a plan. A chart cannot reliably predict an event within a month. Use official dates, direct communication, and real constraints rather than trying to validate symbolic timing with intuition.
Where AI fits in the tradition
It's worth naming the discomfort directly. Some astrologers argue that handing horary to software breaks something — that the craft is the slow human attention, and automating it produces hollow answers. That critique is real and deserves a serious reply.
The reply begins with a distinction: much of horary is a rule system, but judgment is not merely a lookup table. Lilly showed rules, exceptions, context, and worked cases. Software can calculate and expose the rule-bound part with unusual consistency; it should also make its assumptions visible.
What software cannot do is the human work around the chart — helping someone discover what they are actually asking, recognizing what was left unsaid, steadying them through a difficult answer, and knowing when not to judge. An AI-written reading can be lucid and useful; it is still a generated interpretation that should remain accountable to the visible chart.
What iHorary does
On iHorary, casting is free and AI-free. The site computes the chart and lays out the astronomical positions and traditional evidence so you can inspect the work. If you choose the optional written reading, AI helps scope the question and express a reading grounded in that calculated chart; it does not alter the chart itself.
If you are curious what your question looks like as a horary figure, cast the chart and begin with the evidence. That separation — calculation first, interpretation second — is the part worth preserving.