ALGORITHMIC BIAS AND FAIRNESS: CRASH COURSE AI #18

True or False: Algorithmic bias occurs when algorithms are created by machines without human input.

• True

• False

Which of the following is NOT a type of algorithmic bias mentioned in the video?

• Hidden biases in training data

• Lack of representation in training data

• Insufficient diversity in algorithm developers

• Difficulty quantifying certain features in training data

True or False: Positive feedback loops in algorithms can lead to amplification of past biases.

• True

• False

What can happen if an algorithm is trained on data that is biased or discriminatory?

• It will always make accurate predictions.

• It will be unable to recognize patterns.

• It may perpetuate bias and discrimination.

• It will become transparent and unbiased.

How can humans help reduce bias in algorithms?

• By ignoring the recommendations of algorithms.

• By criticizing the recommendations of algorithms without evidence.

• By adjusting algorithms to be more biased.

• By paying attention to algorithmic recommendations and ensuring fairness.

True or False: Transparency in algorithms allows us to understand how and why they make certain recommendations.

• True

• False

What is one potential concern about collecting more data on protected classes?

• It could lead to more accurate algorithms.

• It could violate privacy rights.

• It could eliminate all bias from algorithms.

• It could increase discrimination further.