Treat ideas as hypotheses, generate multiple versions cheaply, and decide by testing them against reality.

Initial ideas are usually wrong and that cheap prototyping changes the economics of exploration. #