Not terrifying. More like boring (in the sense of evoking ennui). I used to work in missile guidance: target identification algorithms. My specialty was kinematic discrimination. People at large (and on this page) have no idea how truly colossal the problem is of devising an artificial mind. Autonomous behavior is relatively simple, so long as it is limited to a few degrees of freedom, and there is a clear guidance algorithm. But I wouldn't step foot into an autonomous automobile---it is just a lucky deathtrap.

My point is that we are a LONG, LONG ways from "artificial intelligence." The stupid robot script-parsers that make cold calls are the best we can do, and it doesn't take too much to "break" their script, as clever and disingenuous and deceitful as they are. Even when we think we've got it down pat, our arrogance and sloppiness is deadly (think 737 MAX).

The story of Frankenstein is a valid warning against thinking we can get there from here. (Not so much that AI is impossible, but that all human creativity is subject to error.)