Quick thought since this lesson has changed a little.
I felt the addition of Joy Buolamwini’s video was wonderful and led to some interesting discussion. However, one should be prepared to address the different viewpoints that inevitably arise, including those to the effect “I don’t see a problem.” (Yes, that was a real comment, by more than one student). Prepare for what you are going to say, but obviously let other students take the reign as they are more often than not very eloquent and insightful in addressing their peers.
I highly suggest a follow-up that is not part of the lesson but that I felt really drove home the message behind Joy’s video. As luck would have it, the code.org team sent out their most recent newsletter the day before I presented this lesson. It contained a link to an Hour of Code Activity called ‘AI for Oceans’. I adapted it and followed up on the discussion by having the students do the machine learning bubbles and watching the videos as a class. It REALLY made the kids understand, in a way that even the ‘I don’t see a problem’ kids could grasp, why representation in CS is important and how bias (both conscious and unconscious) is a problem in the development of artificial intelligence. It’s easy to do, takes less than an hour, and is well worth it. It’s also a jumping off point to more AI lessons, which I supplemented the curriculum with last year.
Hope this helps.