In areas such as transportation, finance, logistics or engineering, the availability of unprecedented computational resources, vast datasets, and novel algorithms is opening pathways to solve problems that were previously intractable due to their large scale and complexity. Material and molecular discovery are hardly an exception. Experimental synthesis and testing are costly and time consuming. High-throughput automated experimental testing is a promising avenue to bring some of these gains onto materials discovery. It is however, very capital intensive and not necessarily more cost-effective than traditional approaches. On the other hand, many key desirable properties of materials can be predicted from simulation,...