A recent discussion regarding a new paper highlights the illusion of understanding in science, noting that researchers often mistake the predictive accuracy of mental models for true comprehension of underlying causal structures. While these simplified models are indispensable, the conversation warns that relying heavily on computational software can constrain how we conceptualize complex problem spaces. Ultimately, the discussion emphasizes that genuine scientific progress requires acknowledging the limitations of these models and avoiding dogmatic adherence when they inevitably fail.