Multiple-choice questions (MCQs) are often seen as a less effective learning tool than open-ended questions. However, MCQs are undeniably easier to grade, making them a practical choice for large-scale assessments. But what if MCQs are just as good for learning, especially when time is limited? A new study explored this question, examining the effectiveness of MCQs versus open-ended questions, and a combination of both, in tutoring lessons focused on advocacy skills. Surprisingly, researchers found no significant difference in learning outcomes between the three question formats. Learners in all groups performed similarly on post-tests, regardless of whether they practiced with MCQs, open-ended questions, or a mix. However, there's a twist: the MCQ-only group spent significantly less time completing the lessons. This suggests MCQs might be the more efficient choice, delivering comparable learning in a shorter timeframe. To further explore efficiency, the researchers used GPT-4 to auto-grade the open-ended responses. The results were promising, with the AI demonstrating a good ability to assess responses, opening up exciting possibilities for automating the grading process. This study challenges traditional assumptions about MCQs and opens up new avenues for designing efficient and effective learning experiences. It also highlights the growing role of AI in education, particularly in automating time-consuming tasks like grading, potentially freeing up educators to focus on what matters most: helping students learn.