Raw LLM Responses

Its not necessarily because the data sets are biased or because the AI is Racist, Or its creators, but rather the data that exists... pretty much world wide tends to favor these categories in most regards. (not an opinion nor am I saying/insinuating anything) But these AI's tend to work off of probability... very basic statistics. That means that the studies/data the AI receives will be averaged between distinct categories. Race, Sex, Age etc... This will mean that an AI targeted at policework... in the US anyway, will tend to disproportionately target Black People over most other groups, given the existing information we have on crime rates tending to be higher among Black communities than most others. (on average) The same goes for Males being targeted more than Females.... same reasoning. Now the AI isn't necessarily racist, sexist or wrong... its working with the data available... and the data isn't necessarily wrong either, but rather it doesn't take into account current attitudes to these issues. Data doesn't give a single toss about you or anything about you... you are a series of numbers on a spreadsheet. (opinion time) The solution to these issues with AI is to remove these divisive categories from the Data, each person in a city is given a number (drivers license, passport, something on their birth certificate etc... That is your identifier. then the relevant information about the person/ their environment/ whatever else is fed to the AI. Race, Sex etc... not needed. The issue is that these things are often intrinsically tied to a person and are difficult to separate. so we are at a loss. Now skin colour and sexuality can be removed but are significant identifiers and statistically indicative of lifestyle, company etc... the sheer amount of information needed to give a policing AI enough information to predict criminals... is more totalitarian than the surveillance in China, which (as far as we are aware) cannot predict crime... yet.