[crawfordcomeaux]

Depth-first search through a concept space doesn't lead to originality as quickly as performing a breadth-first search at each new level of depth.

The shortest path to an idea may require disparate ideas.

I know I'm only rephrasing the headline, but the article's point follows easily from this model of how thoughts are structured in Memory Evolutive Neural Systems.

https://doi.org/10.1016/j.pbiomolbio.2015.07.004

(to read for free, go to https://sci-hub.bz and search for 10.1016/j.pbiomolbio.2015.07.004)

EDIT: To be clear, the point follows easily from the model, which says nothing about what's required to understand the model.