A plain answer to the question "why wouldn't I just ask ChatGPT for a hexagram?" This is not a claim of being more correct. It is about where the hexagram comes from, where the text is quoted from, and what the interpretation is grounded in.
1The hexagram comes from the ritual, not from the model
The yarrow-stalk procedure is carried out as the classical 筮法 division process, so the hexagram falls according to the genuine, uneven probability distribution over the 64 outcomes. A general chatbot asked for a hexagram simply generates one that sounds fitting for how you phrased the question — the answer is shaped by the asking. Here the cast is fixed before any language model is involved, and the model only interprets what the stalks gave.
2The classical text is retrieved, not remembered
The 卦辞 and 爻辞 shown are looked up from a fixed, deterministic store of the 64 hexagrams in classical Chinese, Traditional Chinese, and the 1893 English edition. A chatbot reciting hexagram text from memory misquotes and invents lines; this cannot, because it is not generating that text at all.
3Grounded in real cases, and it will show you them
Interpretation is retrieved from 88 source-verified divination cases from the historical Takashima record — actual Meiji-era readings with their recorded outcomes — not from a generic sense of what an I Ching reading sounds like. Every reading can quote its source.
4One tradition, consistently
This reads in the Takashima Ekidan school specifically. A general assistant blends every divination tradition it has ever read into an average of all of them.
5Your readings persist and can be revisited
Readings are saved, and you can record what actually happened afterwards. A chat window is gone when you close it.
6There are boundaries
When a question signals genuine distress, this does not answer it as a fortune — it points to real support.
Is this different from asking ChatGPT for an I Ching reading?
Yes, in where the hexagram comes from. Here it is cast by the classical yarrow-stalk division before any model is involved, so the outcome is not shaped by how the question was phrased. A chatbot generates a hexagram that fits your wording. The classical text is also retrieved from a fixed store rather than recalled, and the interpretation is built from 88 source-verified historical cases.
Where do the hexagram texts come from?
From a fixed store of the 64 hexagrams holding the classical Chinese original, a Traditional Chinese rendering, and the public-domain 1893 English edition of the Takashima Ekidan. The same lookup returns the same text every time; nothing in that layer is generated.
How is the hexagram chosen?
By the classical procedure: fifty stalks divided, counted off in fours, three changes per line, six lines. Because the procedure is followed rather than approximated, the 64 outcomes are not equally likely — they follow the distribution the method actually produces.
Is the interpretation made up?
The prose is written by a model, but its material is not invented. Hexagram and line texts are retrieved from the fixed store, and the case material comes from 88 source-verified readings in the historical Takashima record. A reading can show the sources it drew on. No claim of predictive accuracy is made.
This service is for cultural and reflective use. It makes no claim of predictive accuracy and should not be used as the basis for any decision.