How do you all feel we can prevent prejudice bias, such as with age, nationality, social class, or gender? Consider the scenario where one of these data fields as an input feature could have predictive power, what would be a good practice to ensure our algorithms are fair?