This dissertation presents the Extended Metacommunication Template, a tool based on a set of guiding questions derived from the theory of Semiotic Engineering. We report the results of a study we conducted to evaluate the tool s impacts on the design process of machine learning systems. By having designers and developers answer a set of questions, the tool aims to help them reflect on their interpretations of the design solution, while allowing them to revisit the presuppositions behind it. We then describe a speculative design study and analyze its results, identifying emergent themes that help us understand how the proposed tool may be used. Among the relevant themes identified are: the reflective practice of design, the designer s focus on their use of language, the process of attributing responsibility to the people involved, the use of the ethical framework provided to them, the bioethical principles, and the ways in which the extension of the template may be used.