Pick one of the questions below and answer in a ~half page response

• What labor issues arise with crowd labor? Do these labor issues affect program correctness? Do they introduce ethical obligations for developers?
• What data quality issues might arise with finely-grained crowdsourced labor? How could you mitigate this. Here's an interesting paper if you're interested in this question.
• How do you break a big problem down into an automated sequence of little ones?
• How can machine learning and crowd-sourcing interact to produce a more robust system? If you're trying to label images of cats, and you have one black box that does it via Mechanical Turk, and another black box that does it with machine learning, how do you decide which to use? Can you leverage them together?

Please submit your response by class time on each Monday. Here are the Journal guidelines and submission instructions.