In data science terms, Resource Geologists are known as Domain Experts; people who have expert technical knowledge of both geological scenarios (ore deposits) and geostatistics. During resource estimation, resource geologists use geoscience data (usually assays, complimented by relevant geological information such as maps or 3-D models) to predict the grade and tonnage of an element(s) within a volume, the region to be mined. Some aspects of the resource estimation workflow can be optimised by replacing manual interpretation with automation. For example, machine learning (ML) can be used to create estimation domains with stationary grade populations, and to represent consistent geological volumes. This work is conventionally done as a manual wireframing task undertaken by resource geologists. However, automated ML tasks can ideally be tuned to make them objective, repeatable, and efficient. This outcome is preferable to solutions which vary based on the user. Here, we briefly summarise the use of ML through time, with a focus on resource estimation, and examine some case studies where it has provided useful input for resource models, and explore what the future of ML may look like. CITATION: Gazley, M F, Hood, S H and Sterk, R, 2019. Machine learning and resource geology, in Proceedings Mining Geology 2019, pp 150153 (The Australasian Institute of Mining and Metallurgy: Melbourne).