I'm not worried so much about "superintelligent" AI, I'm worried about the growing trend of outsourcing human judgement to automated systems. Examples include

I'm not worried so much about "superintelligent" AI, I'm worried about the growing trend of outsourcing human judgement to automated systems. Examples include automatic flagging of content based on misjudged copyright issues, automatic rejection of refund or other claims (especially where all subsequent attempts at resolution also involve non-human decisions), etc. We're not even to the point of superintelligent AI, but some major companies (some which start with G, just as one example) already handle situations incorrectly because of too much dependence on automation.