Part II: Giving Voice to the Social Sciences

第二部分：为社会科学发声

The Future of Reading Comprehension, Where Literature Comes Alive!

阅读理解的未来，让文字生动起来！

Saanvi, a ninth grader in India who attends Khan World School, a full-time online school we started with Arizona State University, was reading The Great Gatsby for class. Her teacher gave her a video assignment to discuss several key questions about the novel, and Saanvi came to one that stumped her. “In The Great Gatsby, why does Jay Gatsby keep looking at the green light?”

Saanvi 是印度一名九年级学生，就读于我们与亚利桑那州立大学共同创办的全职在线学校可汗世界学校。她正在为课堂阅读《了不起的盖茨比》。她的老师给她布置了一个视频作业，讨论关于这部小说的几个关键问题，Saanvi 遇到了一个难题：“在《了不起的盖茨比》中，为什么杰·盖茨比一直盯着绿灯？”

Uncertain of the answer, she did what many industrious students do when they find themselves stuck: she started running a bunch of web searches on the topic. She wasn’t really satisfied with what she found, so she took it a step further.

不确定答案的她，做了许多勤奋学生在遇到困难时会做的事情：她开始在网上搜索这个话题。她对找到的结果并不满意，所以她更进一步。

She turned to Khanmigo . . . and asked to talk to Jay Gatsby himself.

她转向了 Khanmigo……并请求与杰·盖茨比本人对话。

JAY GATSBY: Ah, splendid to meet you, old sport. I’m now Jay Gatsby, the enigmatic millionaire from F. Scott Fitzgerald’s classic.

杰·盖茨比：啊，很高兴见到你，老运动员。我现在是杰·盖茨比，F·斯科特·菲茨杰拉德经典作品中的神秘百万富翁。

SAANVI: Mr. Gatsby, why do you keep staring at the green light?

SAANVI：盖茨比先生，你为什么一直盯着绿灯？

JAY GATSBY: Ah, the green light. It is a symbol of my dreams and desires. You see, it is situated at the end of Daisy Buchanan’s dock, across the bay from my mansion. I gaze at it longingly, as it represents my yearning for the past and my hope to reunite with Daisy, the love of my life.

杰·盖茨比：啊，绿灯。它是我梦想和欲望的象征。你看，它位于黛西·布坎南的码头尽头，在我豪宅对面的海湾。我渴望地盯着它，因为它代表着我对过去的渴望以及我对与我一生挚爱黛西重聚的希望。

“We just started talking,” Saanvi tells me. “There I was, having a great conversation with Jay Gatsby. At one point I felt bad that I was taking up all of Mr. Gatsby’s time, and so I apologized to him! And Gatsby said, ‘Oh, no, I’m not really Jay Gatsby, I’m just an artificial intelligence simulation.’ ”

“我们刚开始聊天，”Saanvi 告诉我。“我在那里，与杰·盖茨比进行了一场精彩的对话。有一刻我觉得不好意思，因为我占用了盖茨比先生的所有时间，所以我向他道歉！然后盖茨比说，‘哦，不，我不是真的杰·盖茨比，我只是一个人工智能模拟。’”

This is exactly the type of interaction that engages our learners and broadens their minds. Today, you can talk to Victor Frankenstein, Hester Prynne, or Odysseus.

这正是吸引我们的学习者并开阔他们思维的互动类型。今天，你可以与维克多·弗兰肯斯坦、赫斯特·普林或奥德修斯对话。

Beyond facilitating reading comprehension, AI can allow learners to immerse themselves in the worlds of the characters in ways that would have seemed like science fiction only a few years ago.

除了促进阅读理解，AI 还可以让学习者以前所未有的方式沉浸在角色的世界中，这在几年前还像是科幻小说。

For educators, these AI writing and comprehension tools help them better understand the ways that their students read and process information. I think we can all agree that reading comprehension is extremely important. It is hard to navigate the world, much less make informed decisions, without it. One could argue that you can’t even begin to write well without first being able to read well. Unfortunately, we are currently in a pretty bad state. Based on a 2020 Gallup analysis of data from the U.S. Department of Education, 54 percent of Americans between the ages of sixteen and seventy-four read below a sixth-grade level.

对于教育者来说，这些 AI 写作和理解工具帮助他们更好地了解学生如何阅读和处理信息。我认为我们都同意，阅读理解极其重要。没有它，很难在世界上立足，更不用说做出明智的决策了。有人可能会说，如果你不能先读得好，你就根本不可能写得好。不幸的是，我们目前的情况相当糟糕。根据 2020 年盖洛普对美国教育部数据的分析，16 到 74 岁之间的美国人中有 54% 阅读水平低于六年级水平。

This is a complex issue, but I have high hopes that AI will make a meaningful dent here.

这是一个复杂的问题，但我对 AI 在这方面有所作为抱有很高的期望。

In most schools, the formal practice of reading comprehension comes from studying books and articles. The evidence of a student’s understanding is usually expressed through some type of essay or in-class discussion. As rich as that experience might be, it is unfortunately hard to standardize and scale. Because of this, traditional passage-based multiple-choice questions are how most students are assessed on high-stakes tests (for example, SAT, ACT, state end-of-year exams).

在大多数学校，阅读理解的正式练习来自于学习书籍和文章。学生理解的证据通常通过某种类型的作文或课堂讨论来表达。尽管这种体验可能非常丰富，但不幸的是，难以标准化和规模化。正因为如此，传统的基于段落的多项选择题是大多数学生在高风险考试（例如，SAT、ACT、州年终考试）中授受评估的方式。

Passage-based multiple-choice questions aren’t inherently bad. In fact, they can be a great way to practice and assess some dimensions of reading. But they are limited in what they can do. Because of the built-in incentives for measuring student success through standardization, educators working with underperforming students tend to emphasize in-class work that mimics these multiple-choice exams, but when they do so, these students get a very narrow exposure to reading (and it isn’t even clear that this approach enhances student performance on those multiple-choice exams). The focus on assessments keeps students from engaging with a more diverse set of texts and modalities. For example, most educators intuitively recognize that allowing students to give free responses and engage in dialogue about a text would likely create deeper readers. They also see that pairing reading comprehension with writing is an ideal way to practice both. Unfortunately, these types of activities are hard to standardize and evaluate at scale.

基于段落的多项选择题本身并不坏。事实上，它们可以成为练习和评估阅读某些方面的好方法。但它们的能力有限。由于通过标准化来衡量学生成功的内在激励措施，与表现不佳的学生合作的教育工作者倾向于强调模仿这些多项选择考试的课堂作业，但这样做时，这些学生接触到的阅读非常狭窄（甚至不清楚这种方法是否能提高学生在这些多项选择题上的表现）。对评估的关注使学生无法接触到更广泛的文本和模式。例如，大多数教育工作者直觉上认识到，允许学生对文本给出自由回答并参与对话可能会培养更深入的读者。他们还看到，将阅读理解与写作相结合是练习两者的理想方式。不幸的是，这些类型的活动很难标准化和大规模评估。

But what if we could create easily accessible, standardized practice and assessment in reading comprehension that is not multiple choice? Imagine if the assignment allows students to give free responses when discussing a text. Well, this is exactly the type of thing that large language models can be good at.

但是，如果我们能够创建易于访问的、标准化的阅读理解练习和评估，而不是选择题呢？想象一下，如果作业允许学生在讨论文本时给出自由回应。这正是大型语言模型擅长的事情。

Instead of answering multiple-choice questions about a text or passage, imagine that students write out the author’s intent behind a choice of words or explore the main idea of the passage (while highlighting those parts of the passage). Picture the AI then asking follow-up questions based on what a student writes. It could ask students to draft a conclusion for an incomplete essay that forces them to understand everything that came before. Based on these interactions, it can then provide feedback on their comprehension to the student and teacher. We are already working on such tools. So far, our team has found that with thoughtful prompting, it can get a large language model like GPT-4 to ask good questions and engage in a meaningful discussion about a topic. The challenge is to ensure that the AI is assessing well and doing it consistently. I believe we will get there.

与其回答关于文本或段落的选择题，不如让学生写出作者在选择词语背后的意图，或探索段落的主旨（同时突出段落中的那些部分）。想象一下，AI 根据学生的写作提出后续问题。它可以要求学生为不完整的文章草拟一个结论，迫使他们理解之前的一切。基于这些互动，它可以向学生和老师提供关于他们理解的反馈。我们已经在开发这样的工具。到目前为止，我们的团队发现，通过深思熟虑的提示，它可以让像 GPT-4 这样的大型语言模型提出好的问题并就某个主题进行有意义的讨论。挑战在于确保 AI 进行良好的评估并且始终如一。我相信我们会做到的。

Let’s take this idea beyond passage-based questions. Imagine if, when reading a book, a student could have a discussion with the AI at the end of every chapter. The AI might ask the student what they think about the book so far, or whether anything was particularly interesting or confusing about the material. It might inquire about major themes or whether the student agrees with a character’s point of view or action. This would all happen through a Socratic dialogue. The AI would give feedback to the student and also report this feedback to the teacher. The student could of course ask any question they like about any aspect of what they are reading. Think about how much more engaging this would be for students than the traditional book report. It is also far richer pedagogically; book reports today already often entail students summarizing the plot while including ideas they pick up off the internet or CliffsNotes.

让我们将这个想法扩展到基于段落的问题之外。想象一下，当阅读一本书时，学生可以在每章结束时与 AI 进行讨论。AI 可能会问学生目前对这本书的看法，或者是否有任何特别有趣或困惑的地方。它可能会询问主要主题，或学生是否同意某个角色的观点或行为。这一切都会通过苏格拉底式对话进行。AI 会向学生提供反馈，并将这些反馈报告给老师。当然，学生可以就他们阅读的任何方面提出任何问题。想象一下，这对学生来说会比传统的读书报告更有吸引力。这在教育上也更丰富；今天的读书报告通常包括学生总结情节，同时包括他们从互联网上或 CliffsNotes 中获得的想法。

We do not need to limit this type of reading comprehension practice to a language arts classroom, as it can extend to any type of textbook or article. Students could even design experiments with a simulation of Marie Curie or co-write a Federalist Paper with a simulation of James Madison or Alexander Hamilton.

我们不需要将这种阅读理解练习限制在语言艺术课堂中，因为它可以扩展到任何类型的教科书或文章。学生甚至可以与模拟的居里夫人一起设计实验，或者与模拟的詹姆斯·麦迪逊或亚历山大·汉密尔顿一起合写一篇《联邦党人文集》。

“点赞有美意，赞赏是鼓励”