Question: Let's say a widespread facial-recognition AI program is found to have been trained using biased data, but the training data hasn't been linked to cases
Let's say a widespread facial-recognition AI program is found to have been trained using biased data, but the training data hasn't been linked to cases of discrimination yet. Should it be taken out of use and rebuilt to be more equitable? Keep in mind that training and re-training these programs requires a lot of resources: time, money, labor, and computing power (which has environmental repercussions). What do you think should be done in this situation
Step by Step Solution
There are 3 Steps involved in it
Get step-by-step solutions from verified subject matter experts
