AI, Responsibility and Exclusion
Great article. Worse, most of AI, isn't, it's statistical optimisation coupled with a surrender of human responsibility. Thus, the jokes about 'computer says no' are coming true.
But, worse still, the 'predictive abilities' will probably be used to exclude and marginalise, in the same way that districts were redlined (qv) in urban USA. And worse, worse, all the 'optimisation' is sub-symbolic so there's no explanations, just judgements and a total inability to deal well with outliers, like the 2008 crash. Note that one of the big hedge funds has decided to use 'AI' to 'manage' its investments: https://www.theguardian.com/technology/2016/dec/22/bridgewater-associates-ai-artificial-intelligence-management
Biting the hand that feeds IT
