“The concept is called ‘reverse inference.’ It’s both neuroscience’s greatest ambition, and the origin of its most frustratingly breathless overstatements.”