|
|
|
|
|
|
by Doingmything123
2407 days ago
|
|
|
I guess the question is how well to we expect AI/ML systems to be able to generalize on a biased subset data. If you are trying to use one of these systems to make hiring decisions but you train it on data from the hires you made, should we really be surprised if biases in the original hiring(education level, gender, race, income, personality) are reflected in the AI's decisions? |
|
|