Status quo bias; System justification

A related problem is that because many AI and non-AI algorithms rely on past data to make predictions, these algorithms can reinforce status quo biases. The algorithm doesn't have to be very advanced or "intelligent" for this to be the effect. If the results are trusted inappropriately, the result can be propagation of harmful situations forward in time. Here's an article discussing a healthcare algorithm that reinforced a racial bias in spending and simultaneously caused funds to be spent very inefficiently: https://lnkd.in/e5_kiHhw 

https://lnkd.in/eAA2iFbn