Scientific Intuition

Preamble

Scientific Intuition develops the judgment before the answer: what feels generative, what deserves a test, and which questions are worth the next experiment.

---

The Signal Is The Question

Execution loops can optimize a task once the task is chosen. This branch moves earlier. The learnable object is the research character that notices which anomaly has heat.

observation -> question -> prediction -> outcome -> intuition mutation -> next question

The evaluator judges question quality: falsifiability, novelty with discipline, downstream usefulness, and whether the model updated its sense of where to look next.

The Link To Idea Engine

This branch feeds 024 Idea Engine because idea quality depends on the taste for problems before it depends on the production of answers. The familiar failure mode is a stream of plausible questions that sound research-shaped and never become sharper after contact with evidence.