Three years ago, James Lloyd and I presented the Automatic Statistician at the joint MSR Cambridge machine learning meeting. The aim of this project is to automate the exploratory analysis and interpretation of data. Although progress has been slow, the project has not been dormant. We have completely refactored the code and extended the scope from time series to include classification, clustering and learning simple graph structures from data. We have built a system to automatically allocate computational resources using Bayesian optimisation. And finally, the ambition of the project has grown to include all aspects of automating machine learning.