By Mike Ananny
Algorithmic "mistakes" are windows into the social and technological forces that create computational systems. They reveal the assumptions that drive system designers, the power that some people have to define success and call out errors, and how far institutions are willing to go to fix failures. Using a recent case of facial detection and remote proctoring, Ananny suggests "seeing like an algorithmic error" as a way to turn seemingly simple quirks and individually felt glitches into shared social consequences that shape social life—that is, into public problems.
