AI 与写作2015labs.foxtype.com
FoxType:用自然语言处理帮你写得更礼貌
Show HN: FoxType – Using NLP to help you write more politely
导读
FoxType 是一个用 NLP 给文本“礼貌程度”打分的在线小工具,作者在 HN 发布后,评论区成了大家花式测试它的地方。对写作者来说,这些测试暴露了一个真问题:礼貌分数不等于得体——堆砌客套话能拿满分,真诚的赞美反而被判“不礼貌”。值得一读,是因为它提醒我们:语气里可被量化的是表面特征,而分寸感最终还得靠人判断。
可以直接用的写作要点
- 01
客套话叠加并不等于更礼貌:把 please、thank you、kindly 堆满一句能刷到 100 分,但真人读起来只会觉得别扭,甚至像在施压。
- 02
标点会改变语气:“Ok let's go” 末尾加一个句号,礼貌分掉了 3 分——短句加句号在聊天语境里常被读成冷淡、想结束对话。
- 03
表面的礼貌措辞可能掩盖真实的攻击性,反过来真诚的赞美(“You are wonderful”)也可能被判不礼貌;写的时候该问“读者会怎么感受”,而不是“这句够不够客气”。
- 04
想检验自己的措辞,可以找同事或朋友当“真人评分器”,把改写的句子念出来看对方反应,比看分数可靠。
HN 评论精选
工具只看表面措辞,测不出真实语气;但标点确实会改变一句话的冷暖。
我用它测了 Key and Peele 短剧《短信误会》里的对话:结果是“Sorry dude, missed your texts. I assumed we'd meet at the bar. Whatever, I don't care.”拿了 71 分(礼貌),“Ok let's go”拿了 76 分,而真正翻脸的“You're the one who's fucking priceless!”只有 6 分。有意思的是,在“Ok let's go”后面加个句号,分数掉了 3 分,变成 73。
chavesn查看英文原文
I used this to check the text messages from Key and Peele's "Text Message Confusion" skit[1]: Keegan: "I've been trying to reach out to you all day, are we on for tonight?" Polite (71 / 100) Jordan: "Sorry dude, missed your texts. I assumed we'd meet at the bar. Whatever, I don't care." Polite (71 / 100) Keegan: "Do you even want to hang out?" Impolite (30 / 100) Jordan: "Like I said, whatever." Neutral (47 / 100) Keegan: "Jesus, you are fucking priceless" Impolite (30 / 100) Jordan: "You're the one who's fucking priceless!" Impolite (6 / 100) Keegan: "You wanna go right now?" Impolite (31 / 100) Jordan: "Ok let's go" Polite (76 / 100)[2] Keegan: "You wanna really do this now?" Impolite (13 / 100) Jordan: "Fuck yeah, let's do it" Impolite (19 / 100) Jordan: "First round's mine" Neutral (43 / 100) [1]: https://www.youtube.com/watch?v=naleynXS7yo [2]: Adding a "." to the end of the "Ok let's go" drops it by three points to 73 / 100
把礼貌词堆到极致就能拿满分,说明这类评分奖励的是表面客套,而不是真正的得体。
我输入了一句疯狂堆砌客套话的请求:“Please would you kindly help if good with you please, thank you kindly, thanks.”,和一位同事一起试了好一会儿,才把它刷到 100 分。
cmccart查看英文原文
>Please would you kindly help if good with you please, thank you kindly, thanks. Took a while, but along with a coworker, we were able to get 100.
真诚的赞美被扣分、阴阳怪气反而得分,说明礼貌评分和真实善意是两回事。
我不是在挑刺,只是觉得试着“打败”这个系统挺有意思。被判“不礼貌”的有:“You are wonderful.”、“I really love it.”、“Your work has really inspired me.”、“Why do I love what you're doing? Because of the following reasons.”;而被判“礼貌”的却是:“Again, please show me your failure, which I love to see.”
JacobJans查看英文原文
This isn't a criticism, per se, but I do find it interesting to try to "beat" the system. Impolite: "You are wonderful." "I really love it." "Your work has really inspired me." "Why do I love what you're doing? Because of the following reasons." Polite: "Again, please show me your failure, which I love to see." Ok. I'm out of time. :)
同类还值得读
我是肯尼亚人:不是我把 ChatGPT 写成了我,而是它写起来像我
I'm Kenyan. I don't write like ChatGPT, ChatGPT writes like me
一位肯尼亚作者反驳“我写得太像 AI”的指责:不是他写得像 ChatGPT,而是 ChatGPT 写得像他——那种平滑、无地域色彩、貌似中立的英语,本来就是殖民教育与精英阶层塑造…
AI 与写作2025 · 799 分 · 504 评4 条写作要点编辑精选使用 AI 助手完成论文写作任务会累积认知债务
Accumulation of cognitive debt when using an AI assistant for essay writing task
这篇论文研究的是:在完成议论文写作任务时使用 AI 助手,是否会让写作者积累"认知债"——也就是把本应由自己承担的思考外包出去,短期省力、长期削弱写作与思考能力。
AI 与写作2025 · 391 分 · 426 评4 条写作要点编辑精选学生使用 ChatGPT 写作指南
A Student's Guide to Writing with ChatGPT
这是 OpenAI 面向学生的一份写作指南,讲的是如何把 ChatGPT 用作写作过程中的辅助工具,而不是替你写完作业的"枪手"。
AI 与写作2024 · 323 分 · 309 评4 条写作要点编辑精选Grammarly,你也背叛了我?
Et Tu, Grammarly?
本次输入里只给了标题和链接,正文和评论都是空的,所以没法据实写出导读。
AI 与写作2025 · 279 分 · 74 评编辑精选用 GPT-3 进行创意小说写作
GPT-3 Creative Fiction
Gwern 的长文系统考察 GPT-3 在创意小说写作上的能力与边界:它能维持人设、写出讽刺腔调,也会整段复述训练数据里的歌词。
AI 与写作2020 · 234 分 · 97 评4 条写作要点编辑精选