Are we ready to let machines do the work and can now concentrate on the science questions ? Have we reached the point where statistics and machine learning are useful tools rather than nuisances and the practice of the few ? I will present four examples where significant progress have been made in these ideas and where the interdisciplinary practices have brought fruitful results. These include, automatic classification, Bayesian parameter inferences, event detection and anomaly detection, and design of surveys and follow ups.