Prompts work best when they retain simplicity and potential rather than over-specification.
The uncarved block, or pu, represents raw potential before form is imposed. In prompt engineering, this teaches that the most effective prompts often contain the least specification—space for the model to respond naturally rather than locked into rigid constraints. An overly detailed prompt, loaded with rules and exceptions, paradoxically produces worse results because it forces the model into artificial postures. A simple, clear instruction that trusts the model's training allows genuine capability to emerge. This mirrors how a Taoist master teaches: not through elaborate instruction but through removing obstacles. The most generative prompts carry openness; they establish direction without dictating every step. This doesn't mean vagueness—clarity of intent remains essential—but it means avoiding the micromanagement impulse. By treating the prompt as pu, potential waiting to unfold rather than a rigid mold, users discover that AI systems often exceed expectations. The wisdom lies in knowing when to specify and when to trust emergence.
Peri can explain this concept, give practical examples, help you decide whether it applies to your situation, or recommend a journey if appropriate.
Explore related journeys or tell Peri what you're working through.