AI 与写作2020gwern.net编辑精选
用 GPT-3 进行创意小说写作
GPT-3 Creative Fiction
导读
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。HN 评论既贴出令人惊讶的生成片段，也指出“背课文”式的抄袭问题，值得写作者一读——它既是一面镜子，照出我们对“文风”的直觉，也提醒我们警惕陈词滥调与无意识的借用。
可以直接用的写作要点
- 01
让两个角色对同一件事采取完全相反的态度（一个亢奋宣讲、一个敷衍应付），幽默与人物性格会自然浮现，不必靠旁白解释。
- 02
把抽象情绪拆成具体的排比清单（免税、独裁、调情……），空泛的“自由”才落地成读者能感到的东西。
- 03
让克制的一方用极短的回应（“Sure, I guess.”）去衬托另一方的长篇独白，节奏对比本身就是刻画。
- 04
凡是被模型整段吐出的句子，往往是训练语料里最“顺口”的表达——写作者也该自查：这句是不是我借来的现成话？
HN 评论精选
对话两极反差+清单式排比，是制造黑色幽默和人物性格的高效手法。
它写出的《小马宝莉》同人小说有种让人不安的连贯与机锋：刚死去的教授兴奋地宣布自己从此不必再担心偶数、概率、战争和调情，获得“自由”；而公主只能不断敷衍“是啊”“大概吧”“你想让我说什么呢”。
mintplant查看英文原文
It writes uncomfortably coherent and witty fatalist My Little Pony fanfiction: > “I’m Princess Celestia, of Equestria.” the alicorn said, “It’s nice to meet you. And I’m very sorry for your untimely demise.” > “Heh. You’re telling me.” the Professor laughed, shaking his head. “I’m dead. I’m… dead. Do you know what that means?” > “I have an inkling.” Princess Celestia said, wincing. > “It means I don’t have to worry about even numbers anymore.” he said, nodding. “I don’t have to worry about probabilities anymore! I don’t have to worry about the upcoming war anymore! I don’t have to worry about anypony anymore!” he spread his hooves out at his sides, “I don’t have to do anything anymore! It’s beautiful!” > “Yes, yes, I’m sure it is.” Princess Celestia said, nodding along. > “Nothing matters now that I’m dead!” he cheered, “It’s great! I’m free! This is what it feels like to be liberated!” > “I can see that.” she said. > “I’m free from death!” he continued, “I’m free from taxes! I’m free from dictatorships! I’m free from flirting!” he laughed, “I’m free from everything! Isn’t it great?” > “Sure, I guess.” Princess Celestia shrugged. > “What do you think about all of this, Princess?” he asked. > “About what?” she replied. > “Well, I just thought you might have something to say about it all.” he told her. > “What would you like me to say?” she responded. > “Do you think death is awesome?” he questioned. > “Well… no.” she shook her head, “But I think the passage to it is.” > “Oh.” he said, “I see what you mean.”
模型会原样 regurgitate 训练语料中的名句，写作者同样要警惕无意识的陈词与借用。
在文中“苏斯博士《 Oh, the Places You'll Go 》”那一节，它直接照搬了披头士的歌词：“There's nothing you can know that isn't known…”
mej10查看英文原文
Straight up plagiarizes The Beatles here: https://www.gwern.net/GPT-3#dr.-seuss-oh-the-places-youll-go " There’s nothing you can know that isn’t known. Nothing you can see that isn’t shown. Nowhere you can be that isn’t where you’re meant to be. "
把时间维度纳入写作素材，能让文字抓住“当下感”而非泛泛而谈。
Carmack 提出：GPT-3 这类大模型把十年互联网文本当作独立同分布数据来训练；如果加入时间信息，它们或许能开始写出“明天的 reddit 或 twitter 热点”。
Reedx查看英文原文
Carmack posted (yesterday) an interesting thought on models like GPT-3: "Big AI models like GPT-3 train on massive internet text dumps, but the data is assumed to be independent and identically distributed. Incorporating time information for a decade of data might allow them to start writing tomorrow's reddit or twitter trends." https://twitter.com/ID_AA_Carmack/status/1278840413919551488
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AI 垃圾文、怀疑与写作的反击
AI slop, suspicion, and writing back
这篇短文讨论的是：当 AI 生成的低质内容（slop）充斥网络后，读者开始对一切文字先抱怀疑态度，默认「这可能又是 AI 写的」。
AI 与写作2025 · 231 分 · 175 评4 条写作要点编辑精选