Postphenomenological studies have explored technological mediation between the human body and the world by analysing the bodily experience of the world. Applying this analytical perspective to predictive technology requires some expansions because humans cannot directly experience the future world. I conceptualize pre-spectival focus, which refers to how human attention is directed to the making-future-present process, and which features or aspects of its process are foregrounded or backgrounded. Through the concept of pre-spectival focus and actor-network theory (ANT), this article examines the case of System for the Prediction of Environmental Emergency Dose Information (SPEEDI), a Japanese technology used to simulate the atmospheric dispersion of radionuclides released from nuclear reactors. SPEEDI provides prediction maps representing radiological consequences and was expected to support evacuation decisions during nuclear emergencies. However, this was not the case with the Fukushima disaster, which led to a socio-technical controversy regarding SPEEDI’s usage. Based on bibliographic surveys and several interviews, I encapsulate four multistable uses of SPEEDI: prediction as supporting advice, prediction as a tool for evacuation drills, prediction as self-protection, and prediction as a source of misunderstanding. Relevant actors perceive the predictions of a nuclear disaster in each stability depending on the diversity of their pre-spectival foci, which is also related to the forms of life nourished through their professional and daily lives. A distinct rivalry can be observed between the two actor-networks around nuclear emergency management in which SPEEDI is differently enrolled: the social control network and self-determination network. In the former, the residents are constituted as passive selves who obediently follow governmental instructions; in the latter, residents are included as autonomous subjects who can actively decide protective actions. Moreover, I discuss future postphenomenology–ANT studies on predictive technologies based on these analyses.