When algorithms mess up, the nearest human gets the blame

A look at historical case studies shows us how we handle the liability of automated systems.

from New on MIT Technology Review http://bit.ly/2Mbv9Sv
>