Ethics and Data Science has two important virtues of being free and short, which make it a decent starting place for a conversation about ethics and data science. However, it doesn't do much to advance the conversation beyond hoary tropes to "do better" with caring for user data.

The basic premise is that programming ethics is more than a code or an oath, it's a daily practice that can made explicit by checklists to question the assumptions going into your program, and "five Cs" to follow, in treating customer data as your own personal data.

Getting ethics right is important. Facebook's Cambridge Analytica related scandals are only the tip of the data iceberg. But I'm not sure that 20th century ideals of informed consent have much to say about the sheer combinatorial velocity of data in the 21st century.