GEUL — A semantically-aligned artificial language for unambiguous communication between humans and AI
Natural language evolved for human communication. Ambiguity, redundancy, and implication are strengths for humans but causes of hallucination for AI. Neither programming languages nor existing semantic frameworks are the answer. A new artificial language satisfying six conditions simultaneously is needed.
Hallucination is not an LLM bug — it is a structural inevitability of natural language. Four flaws — ambiguity, absent sources, absent confidence, absent time — make it unfixable by scaling alone.
Programming languages describe procedures. They cannot describe the world. JSON provides structure but no meaning. Even LISP merely borrows syntax.
Rearranging embedding vectors breaks the model. Avoiding breakage means rebuilding the model from scratch. What we need is not transparency inside the black box, but a transparent layer outside it.
Artificial languages for humans failed — artificial languages for AI are different
GEUL does not reject Wikidata. It transforms the classification system and frequency statistics of 100 million entities into SIDX codebooks. Grammar is built on top of a dictionary.
Building a verb system from scratch means gaps, arbitrary choices, and no justification. WordNet is a 40-year lexical database of 13,767 verb synsets built by linguists. We borrow the dictionary and build the grammar on top.