Today's children grow up in an environment that is increasingly characterized by digital voice assistants (DVAs), such as Alexa, Siri, or the Google Assistant. This paper argues that any attempt to investigate children's interactions with, and perceptions of, DVAs should be based on the theoretical grounds of an ontological framework that considers children's genuine understanding of what it means to be human and what it means to be a machine. Based on focus groups and a gamified data collection design, our empirical inquiry applied qualitative methods to explore primary school children's (n = 27, age range: 6-10 years, average age: 8.6 years) open interactions with DVAs. In particular, our focus was on how DVAs were embedded in children's general ontological belief system, and how children interpreted certain aspects of DVAs' interactive capabilities as being genuinely humanoid or non-humanoid. On the one hand, our findings suggest that children's interactions with DVAs might be more an end in itself than a means to an end, meaning that children primarily interact with DVAs for the sake of engaging excitement instead of using the devices' utilitarian functionalities. On the other hand, we found that children in our sample held firm ontological beliefs about the distinct nature of humans and machines, whilst interpreting certain aspects of DVAs' interactive capabilities as being genuinely humanoid (e.g., non-responsiveness, delayed responses, inaccuracy) and non-humanoid (e.g., permanent responsiveness, promptness, accuracy, limited conversational capacities, lack of common sense, standardized responses) at the same time.