Museums are important spaces for free-choice learning. However, historical objects are often emphasized without explicitly exploring the ongoing relevance of a domain, particularly when addressing emergent technologies such as GenAI. In this paper, we propose a two-layer cooperative game design that combines active exhibit exploration through puzzles with an NPC serving as a Bot of Conviction to promote reflective dialogue as a means of transformation. Playtesting sessions with 10 participants suggest that the puzzle component effectively supported learning about the history of Computer Science (specifically, Human-Computer Interaction), while the collaborative setup was widely described as enjoyable and engaging. Although some participants reported novel insights, others found that the dialogue echoed and helped clarify preexisting reflections, underscoring both the topical relevance of GenAI and the potential of narrative-driven design to support reflective learning.