On Monday, every student in Grade Eight received the same reading recommendation: a fantasy series about a boy with a flaming sword. On Tuesday, the app recommended it again. By Wednesday, Aisha had heard the title so many times that it sounded like a rule. The library's new recommendation system was supposed to help students find books they loved.
星期一，八年级每个学生都收到同一条阅读推荐：一套关于持火焰剑少年的奇幻小说。星期二，应用又推荐了它。到星期三，艾莎听到这个书名太多次，它简直像一条规定。图书馆的新推荐系统本应帮助学生找到他们喜爱的书。
Aisha managed the student book board, a cork display where readers pinned handwritten reviews. She noticed that the app's list rarely included poetry, history, or stories set outside familiar cities. When she asked the librarian, Ms. Chen shrugged. “The system learns from what students borrow,” she said. “And most students borrow what everyone else is reading.”
艾莎负责学生图书板，那是一块软木板，读者可以把手写书评钉在上面。她发现应用的清单很少包括诗歌、历史，或发生在陌生城市的故事。她问图书管理员陈老师，陈老师耸耸肩。“系统根据学生借什么来学习，”她说，“而大多数学生借的是别人都在读的书。”
Aisha thought about her own reading. She had chosen the flaming sword series because a friend had chosen it first. The app was not forcing anyone; it was simply repeating the loudest choice. That felt different from a mistake and more difficult to fix. She took a blank sheet of paper and began listing books the app had never mentioned.
艾莎想到自己的阅读。她选那套火焰剑小说，是因为朋友先选了它。应用没有强迫任何人，它只是重复最响亮的选择。这感觉不像一个错误，却更难修正。她拿出一张白纸，开始列出应用从未提到的书。
Aisha and her friend Omar spent a week checking the library shelves. They found translated novels, slim essay collections, and a memoir by a local nurse. Many had not been borrowed in months. Some were excellent; some were dull. The point was not that every forgotten book deserved attention, but that the app had stopped looking.
艾莎和朋友奥马尔花了一周检查图书馆书架。他们找到翻译小说、薄薄的随笔集，还有一位本地护士的回忆录。许多书几个月没人借过。有些很精彩，有些很枯燥。重点不是每本被遗忘的书都值得关注，而是应用已经停止寻找。
They asked students to fill out cards: the last book you loved, and why. The answers were surprising. A quiet boy loved a bicycle repair manual. A basketball player recommended a novel about grief. None appeared on the app's front page, yet each had a reader who could explain its worth.
他们请学生填写卡片：你最近喜欢的一本书，以及原因。答案出人意料。一个安静的男孩喜欢一本自行车维修手册，一名篮球运动员推荐了一本关于悲伤的小说。它们都没出现在应用首页，却各有一个能解释其价值的读者。
Ms. Chen helped them build a display called “The Missing Shelves.” They did not remove the popular series. They placed it beside the others with a sign: “If you liked the flaming sword, try one of these.” The first day, three students took a chance; the second, seven. The app kept recommending its favorite, but the library had begun to answer back.
陈老师帮他们做了一个叫“失踪的书架”的展示。他们没有撤掉那套热门小说，而是把它放在其他书旁边，并挂上牌子：“如果你喜欢火焰剑，试试这些。”第一天有三个学生愿意冒险，第二天七个。应用仍在推荐它的最爱，但图书馆已经开始回应。
At the end of the month, Aisha was invited to speak to the app's company. She expected a formal meeting, but the designer, a young woman named Ines, listened with a notebook on her knee. Aisha explained that a recommendation was not neutral; it made some books visible and others invisible. Ines wrote that down.
月底，艾莎受邀与开发应用的公司交流。她以为会是一场正式会议，但设计师是一位叫伊内斯的年轻女性，她把笔记本放在膝上认真听。艾莎解释说，推荐并不中立；它让一些书被看见，也让另一些书隐形。伊内斯把这句话记了下来。
The company could not change everything at once. Instead, it added a question to the app: “What did you love about this book?” Students could choose humor, mystery, or a character who felt like a friend. The answers gave the system more than a popularity count. It was a small change, but it made the list less flat.
公司无法一次改变所有事情。于是，它在应用里加了一个问题：“你喜欢这本书的什么？”学生可以选择幽默、悬疑，或“一个像朋友的角色”。这些回答让系统得到的不只是流行度数字。改变很小，却让清单不再那么扁平。
Aisha did not expect the app to become wise. She still preferred the cork board, with its crooked pins and handwritten arguments. Yet she understood that tools could be questioned without being abandoned. When the flaming sword next appeared at the top, she smiled and pinned a poetry collection beside it. “Room for one more,” she wrote.
艾莎并不指望应用变得明智。她仍更喜欢那块软木板，上面有歪斜的图钉和手写的争辩。但她也明白，工具可以被质疑，而不必被抛弃。当火焰剑下一次出现在顶端时，她笑了笑，在旁边钉上一本诗集，写下：“再留一个位置。”