Anyone working in AI should read Lorraine Daston’s Rules: A Short History of What We Live By. It’s a history of rules and why rules are important. There are different levels of what a rule is. The kind we default to now is algorithmic. When we make a rule inside an organization, for example writing a vacation policy, we want it algorithmic: explicit enough that anyone can follow it step by step. Rules used to be what Daston calls "thick." They came with examples, exceptions, and the…