Humanities research with computing is frequently associated with three approaches to technologies: building infrastructure, designing tools, and developing techniques. The infrastructural approach is common among some libraries and labs, for example, where "infrastructure" implies not only equipment, platforms, and collections but also where and how they are housed and supported (Canada Foundation for Innovation 2008, 7). Tools, meanwhile, are usually designed and crafted with infrastructure. They turn "this" into "that": from input to output, data to visualization, source code to browser content (Fuller 2005, 85). Techniques are then partly automated by tools. Aspects of a given process performed manually may become a procedure run by machines (Hayles 2010; Chun 2014). Although these three approaches are important to humanities computing, today they face numerous challenges, which are likely all too familiar to readers of this handbook.Among those challenges is technical expertise. Undergraduate and graduate students as well as humanities staff and faculty are rarely trained in areas such as computer programming and artificial intelligence (AI). Developing this expertise is no small task, especially when it is combined with academic studies of history, culture, language, or literature. Software appears in the meantime to grow and obsolesce rapidly. Just as someone finally learns the ins and outs of a platform, they may be asked-or required-to switch to another one. Computing in the humanities thus brings with it various justifiable concerns, if not anxieties, about the perceived obligation to "keep up" with the pace of the (mostly privatized) technology sector (Fitzpatrick 2012). Alongside this obligation comes the related challenge of maintenance. Infrastructure and tools demand routine attention, even when platforms are automagically updated. Maintenance is expensive, too. If researchers are fortunate to acquire grant funding to build infrastructure or design a tool, then they must also consider the near future of their projects. What will be the state of this humanities platform in ten years? Who will be using it, what will they want or need, who will steward it from here to there, and at what cost? Such issues are labyrinthine in that their trajectories are incredibly difficult to predict. They are also complex from the labor perspective, where precarity is now the default state for academics who are increasingly spread thin yet expected to do more and more with less and less.As we confront these challenges-one of us (Julie) a PhD student in science and technology studies, and the other (Jentery) an associate professor of English-we are experimenting with another approach to computing in the humanities, namely autoethnography, which is by no means new to the academy. Carolyn Ellis and Arthur P. Bochner provide a capacious but compelling definition of autoethnography, and we adopt it for the purposes of this chapter: "an autobiographical genre of writing and research that displays multiple layers of conscious...