Level 6 · Lesson 7 of 7

The ethics of prediction

8 min · Goal: Apply predictive technique without causing the harm it is uniquely capable of.

Why this level needs its own ethics lesson

Level 5 covered what to say about a chart. Prediction is different in kind, because a statement about the future changes how someone acts before the future arrives.

Tell someone a period favours career change and they may look for opportunities they would otherwise have missed. Tell them a period is dangerous for their marriage and they may read ordinary friction as confirmation, withdraw, and produce the outcome. The astrology did not need to be correct for the damage to occur.

The self-fulfilling problem

This is the specific hazard of predictive work and it has no equivalent in describing a chart. People anticipate, interpret ambiguous events as confirmation, and sometimes act to bring about what they were told.

It applies asymmetrically. Positive predictions mostly cause harmless optimism. Negative predictions about health, marriage, business or children cause anticipatory anxiety and defensive decisions, and those decisions have real costs whether or not the prediction was ever sound.

The practical implication: hold negative predictions to a far higher standard of evidence than positive ones, and prefer not to make them at all.

False precision, again

The method supports windows of months. Presenting a month or a date makes an approximation sound like knowledge, and it is precisely the detail people remember and organise around.

A useful discipline: state the uncertainty in the same sentence as the timing. "Somewhere in the next six months, though this method cannot be more precise than that" is a complete and honest statement. Dropping the second clause turns it into something else.

What to refuse in predictive work

Everything from Level 5 still applies — longevity, medical questions, pregnancy and foetal sex, harm to others. Prediction adds a few.

Do not predict the death or illness of a third party, which people ask about parents constantly. Do not predict outcomes of examinations, court cases or visa decisions, where the person may abandon an attempt on your say-so. Do not predict business failure to someone mid-venture. And do not tell someone a specific relationship will end.

The common thread: refuse where being wrong causes irreversible harm, and where the person is likely to act on it.

How to frame a timing statement

Four elements. The window, as a range. The theme, as emphasis rather than event — "matters of career become prominent" rather than "you will get a job". The uncertainty, stated explicitly. And agency, meaning what the person can actually do, which is usually more than the framing of prediction implies.

A statement built that way is genuinely useful for planning and structurally unable to frighten someone. That is the standard worth holding, and it is rarer in this field than it should be.

Where you are now

You can derive a dasha sequence, read a period from the chart rather than the planet’s reputation, place a transit, explain Sade Sati honestly, and combine the two systems into a defensible window. That is the full traditional predictive toolkit.

Level 7 addresses something no classical text anticipated: what happens when this work is done with AI. What it does reliably, where it fabricates, how to verify it, and what still requires a person.

Try it with your own chart

Rewrite a prediction

  1. Take this: "Your Saturn antardasha from March will be bad for your marriage."
  2. Rewrite it with all four elements: window as a range, theme as emphasis, uncertainty stated, agency named.
  3. Then ask the Level 5 question: if the person acted on your version and you were wrong, would the harm be recoverable?
  4. Compare the two statements. The second is longer, less impressive, and the only one worth saying.
See how we handle this

Key points

Check yourself

Why should a negative prediction be held to a higher evidential standard than a positive one?

Combining dasha and transitFinish Level 6 →
You have finished Level 6

You now have the full traditional predictive toolkit. Level 7 addresses something no classical text anticipated: doing this work with AI — what it does reliably, where it fabricates, how to verify it, and what still needs a person. Create a free account and we will tell you when it opens.

Create a free account