返回资讯中心
外部精选
软件工程
#开发者实时流

AI Use in Publishing

I’m rather surprised to find that Wired is the magazine which takes such a knee-jerk anti-AI line. In a piece by Adam Morgan, entitled, Book Publishers Are Quietly Using More AI. Staff Are Revolting (link via LitHub) th…

Hacker News 最新Tomte2026/10/22 · 23 天前8 分钟阅读

以下正文同步自 Hacker News 最新,版权归原站所有,已转换为易读排版。

I’m rather surprised to find that Wired is the magazine which takes such a knee-jerk anti-AI line. In a piece by Adam Morgan, entitled, Book Publishers Are Quietly Using More AI. Staff Are Revolting (link via LitHub) they reveal that “behind the scenes, at least three of the Big Five US book publishers—HarperCollins, Simon & Schuster, and Hachette —have been quietly incorporating AI tools into their publishing processes without public disclosures or author consent.”*

Author consent?! Did we fail to consult our authors when we moved from typewriters to Macs? When we changed our banker from Irving Trust to Chase? When we changed the central heating system from oil to gas? You bet we did. It’s none of their business. You’ve got to be pretty self-righteously worked up to think it scandalous that “publishers are now using LLMs like Claude and ChatGPT to email literary agents, write publicity and back-cover copy, design cover art, and generate marketing materials”. They’d be irresponsible if they weren’t! Authors write the books; publishers contract to make and market them.

The Wired article is based on “interviews with more than two dozen workers, all of whom spoke under the condition of anonymity”. So Wired has obviously uncovered a huge conspiracy of “revolting staff”! Junior staff are being overworked and presumably not being employed in the first place. But, remember, junior staff have always felt overworked. They were/are no doubt absolutely right, but that’s publishing for you. You’ve got to love it: those who don’t quickly leave it.

Is there any hypocrisy lurking in a publisher’s willingness to use AI internally but refusing to countenance its use by authors? I think not. It can still be argued that if you buy a book claiming to be written by Writer-123, there’s a reasonable expectation that it’ll be a book written by Writer-123. When you pick up the book in the bookstore, you don’t imagine that the blurb on the back cover was written by the author, or by some particular publishing assistant. If it does its job, do you care if the copy was written by Alice, or Jimmy or by Claude — even if we assume you could tell the difference, which of course you couldn’t? Surely we will get used to the idea that AI can be a help in writing a book; that it’s not a complete subversion of the ethics of writing. Can’t wait.

The one area in which publishers’ use of AI might be criticized could be in its environmental impacts. A lot of electricity is eaten up by every AI effort, and I heard on the radio the other day that using AI to write an email consumes the equivalent of a bottle of water. Battles over the construction of data centers is the issue du jour in this Fall’s elections. AI’s voracious appetite for capital is driving the stock market these days. If AI uses so much water we need to buckle down (as we need to anyway) and set up more desalination and purification plants. And isn’t electricity eventually going to be free — the sun just gets on with sending it to us all we have to do is capture the light? Changing the name of AI to SI isn’t going to get people to stop hating and fearing AI. The bad bits need to be regulated; the good bits need to be used.

正文由 探How 从来源站点 RSS 同步,内容未经改写;遇到排版缺失或需要图片、视频时请以原文为准。

打开原文
去社区讨论 →