Widespread misperception of phenomena only "matters", socially speaking, when the misperception has real-world consequences.
For any number of things, perception=reality, because "reality" is a contingent construct.
Conversation
Thinking about this as I read a section on how machine learning has "replaced" experts in a number of fields where the purported gains of machine learning are 100% fraudulent.
1
1
It doesn't really matter if it has or not, in the context of the social status of expertise in said fields - it will have the exact same social impact as if it did.
Perhaps more.
Of course, it will matter in the sense that any field which starts using machine learning instead of human experts is likely to suffer an abject drop in performance, if said experts are *real* and not themselves part of a fraudulent field.
But such things are difficult to track.