> When you realize that an algorithm is just a process to be performed, what is new here other than using a computer and a large amount of data? This is precis

> When you realize that an algorithm is just a process to be performed, what is new here other than using a computer and a large amount of data? This is precisely the problem. If the algorithm is dependent on a certain process, this process should be vetted for potential biases - big data or not.