“AI” is a term that crops up with increasing frequency as automated systems perform functions once squarely in the realm of human thought and control.
An initialization of “artificial intelligence” “AI” dates back to the mid 20th century. By the 1970s, the initialization was established beyond computer science circles, even though the technology was not.
Philip K. Dick, in his novel VALIS ,from 1981, throws the term around with little explanation. Granting that he was a science fiction writer addressing an audience more likely to be familiar with the term, it’s still sort of startling to see AI casually discussed during an era in which most cars still sported carburetors and ATMs were considered pretty hot stuff.
From this perspective, the AI that determines what shows up in your Facebook feed seems all but inevitable, a promise fulfilled, but also a trite use of a seemingly miraculous technology.
The fact that Facebook has had to employ actual humans to cull truly objectionable material, though, pulls the curtain on the AI wizard: perhaps this glib term is covering how artificial the intelligence really is. Or maybe it makes opaque some unsettling truths: we have a good idea of what “artificial” means, but maybe not so deep an understanding of “intelligence.”
As with many other initializations and acronyms, the shorthand implies a certainty that does not really exist.
AI projects the veneer of the official onto the rickety structure of the ad hoc.