#hw#cs390#notes Avoiding bias is important for credit card loans, face identification, etc.
Recently with the rise of smart phones and massive amounts of data, machine learning algorithms have become more powerful.
In the US AI is developed for commercial purposes (targeted ads). Facial recognition used in apartment buildings — many tenants are unhappy with it. Brings up the idea that poorer places get the tech first to ‘test’ while rights are being violated, then go to affluent places last.
Amazon tried to use AI to sort resumes, it ended up being biased towards women.
Tenure & Firing Model for teachers are used by schools, but are not much better than ‘random’. Teachers have been discriminated against via these algorithms. Due process laws violated upon firing by ‘algorithm’.
In court a ‘risk assessment’ algorithm, misclassifies black people higher, and white people lower as reoffenders.
Social credit score even impacts how much you ‘trust’ someone in China. Messing up your score can ruin how you are socially.
AI is being used behind the scenes, and it is a real societal issue if AI is used to ‘nudge’ us towards certain forms of thought and action.