Companies typically pitch their family history-supporting products' nifty features, but rarely discuss the underlying models the software uses to store your family history. Those underlying models represent how the software can think about the work on which you and it are going to collaborate; understanding their limitations can help you avoid frustration and guard against your software's limited worldview rubbing off accidentally on you. In this talk I'll peel back the digital skin, discussing how to think about how computers think generally and how to recognize how their limitations are causing them to do surprising or annoying things. I'll also discuss some of the current limitations of most family history programs, together with which of those limitations are likely to be lifted soon and which might be here for the long haul. Coming in to this talk you don't need to know anything about computers, but you should be willing to think about how they think.