Machine learning - just how predictable are we?

Algorithms are already being used to help determine who's approved for a loan, who is the best candidate for a job and which criminal is least likely to reoffend. But, how reliable are they… and what ethical considerations should be applied when using data and algorithms to target consumers? In 2015, a study at MIT suggested that an algorithm could predict someone's behaviour faster and more reliably than humans can. The Data Science Machine, created by a master's student in computer science, was able to derive predictive models from raw data automatically – without human involvement. It's fairly common for machines to analyse data, but humans are typically required to choose which data points are relevant for analysis.