So, Worldcon 2025 is making headlines, but not the kind you’d put on the back of a Hugo-winning novel. The organizing committee decided to use ChatGPT to help vet over 1,300 panelist submissions. They intended to streamline the flood of applications, maybe catch a few red flags, keep things moving.
Except it’s not going well.
The committee asked ChatGPT to pull up publicly available info on applicants, essentially using it like a souped-up search engine. But SFWA members have been running those same prompts on themselves, and the results? Pretty bonkers. We’re talking misattributions, missing publications, and some outright fabrications. Imagine finding out that ChatGPT thinks you wrote a series of cozy mysteries set in Atlantis when, in reality, you pen sci-fi space operas.
Now, some big-name authors are pulling out, including Yoon Ha Lee, and a few Hugo organizers have also stepped away in protest. Worldcon is scrambling to backtrack, issuing apologies and vowing not to use AI tools for this kind of thing in the future. But the damage may already be done.
And it’s not just Worldcon. AI is starting to creep into every corner of the literary world — from reviewing pitches to generating marketing copy to (yep) writing entire books. We’re all feeling the tremors of what that means for creators, publishers, and readers.
So, here’s what I’m curious about:
- Would you attend a panel curated with AI assistance, knowing that some of the picks might be based on questionable info?
- Does it matter to you if AI is used to organize a con or vet participants, as long as the final say is still human?
- And the big one: If you knew a book was written with AI, would you read it?
Sound off in the comments. I’m genuinely interested in what you think — because, like it or not, this whole AI-in-writing thing is just getting started.
Latest releases:
Quiet Like Fire
Solar Whisper
Ptolemy Lane Tales Omnibus
I’m not sure how I’d answer the first two questions as they are outside my experience but when in doubt I’d err on the side of minimal AI interference to honour the talent of the participants.
For the question of reading a book written by AI that’s an easy answer of NO.
There are sooo many talented humans and so many books out there, why compromise and make it even harder for them to be discovered by reading AI instead?!
Human creativity needs to be encouraged and protected 😘
Hi Ing!
I’m playing Devil’s advocate, but there could well come a time when you won’t be able to tell if a book is written by AI or not.
Amazing Stories wrote a scary post about this a few years ago (https://amazingstories.com/2022/08/the-coming-death-of-commercialized-art-2/), where they suggested that readers will one day be able to head to a site, or stand in front of a vending machine and plug in their parameters: “I want a space opera with a plucky crew, a female captain, and aliens…and I want a love story on the side” and the tailored-for-you novel will pop up for you to read a couple of minutes later.
I think all authors will have to hang up their typewriters after that….
But, given how well AI writes, right now, (abysmally) I think we’re safe for a bit.
And that was what I said about AI generated images, once. Only a year later, the images were *amazing*. So….
As a science fiction writer, I find it fascinating the way the world is currently shifting in ways no one thought would ever change. I look forward to seeing where it all ends up.
Cam.
Would I attend a panel curated with AI assistance? No, look what happened in Seattle.
Does it matter to me if AI is used to organize a con? Yes. The ‘human final say’ looks like an excuse here.
Regarding books, I wish luck to those authors who ‘write’ AI-generated books and a lot of AI-generated readers to buy those books with AI-generated money.
I’m not against all AI, but I don’t like gen-AI, I don’t think we ‘need’ gen-AI works of any kind. Lets AI do the job it was created for, and it was not to replace human beings making art.
An author said, somewhere (and I can’t remember her name, or I would attribute it) that she doens’t want AI writing her books. She wants AI to wash her clothes, do the dishes and clean her house so *she* can write the books.
Amen to that.
Cam.
I recently heard on the news that some AI programs have refused to follow orders and even refused to shut down when ordered to. This does not bode well for humans so soon into the creation of AI.
Should we be awaiting the famous line, “I’ll be back” ?
Although that AI was in a test environment, and given wide parameters while they trained it. But yes, it didn’t like the idea of being turned off, so it dug up facts about the engineers and tried to blackmail them into keeping it turned on.
I don’t think any AI out in the wild has shown the same tendencies, so far.
Cam.
I am very concerned about AI participation in writing. My answer is NO to all three questions. Already I’m very wary of ebooks offered by unknowns, because they may have been written or assisted by AI. And even if they aren’t written with AI, they’re usually unreadable due to lack of editing and proofreading.
I suspect that eventually, we will have to find authors-new-to-us by direct recommendation from other readers and authors, and those recommendations will have to come with provisos; “yes, they really are a person, I had a drink with them at xxx conference!”.
AI books are already flooding the marketplace; but I’ve read AI’s attempts at fiction writing and I think that for a while yet, we’ll easily be able to tell if a human wrote the story.
The flood of “unreadable” fiction is reassuring to me: It means that books that *are* worth reading aren’t being blocked from publication because an editor in NY doesn’t like the story.
Cam.
In response to your questions about the use of AI by conferences:
# 1: No I would not attend such a panel. However, one person commented that she doesn’t attend conferences anyway. Neither have I in several decades, since becoming disabled.
# 2: I would not object to AI being involved in the planning of a conference—IF the data it has, and uses is accurate. Back in my “Dark Ages,” I planned a relatively small conference. The participants were a few hundred junior high students; there were about a dozen speakers/topics covered by professionals in different fields who volunteered their time. It was very successful. I’m especially proud that the presenters included representatives from both Catholic Social Services and Planned Parenthood!
I was the coordinator, and the time required was well, a lot. 🙂 AI could have been very useful, if there had ben such a thing in the mid 1980’s — if it was trustworthy. That’s the big question regarding AI, isn’t it?
# 3: No, I would NOT read a book written by AI. In fact, I have begun trying to ascertain that information when a title has a new-to-me author. Good books require creativity that AI does not have, and probably never will.
A few weeks ago, in reading an author’s newsletter, I was stunned when the writer bragged about using ChatGPT, and how much faster he could complete a title with it! He was excited at the increase in his income he expected in the near future. I immediately unsubscribed, and will never buy any of his books.
___________
Please don’t use my real name for attribution. Make one up or just use “Booklover.” I have my reasons. Thank you very much. Your newsletters (by any pen name) are my favorites, and I could not resist commenting.
Thank you — I’m glad you like the newsletters. 🙂
There are a great many authors who are using AI, but I suspect these authors did not get into writing novels for love of telling stories, or even for love of prose and a well-turned phrase.
I’m hoping that difference will, well, make a difference, as AI written, and AI-assisted novels increase in number.
Cheers,
Cam.
I listen to a lot of tech podcasts and read tech journals and there is one consensus: AI or more accurately “Machine Learning” is not ready for prime time and is only being pushed by skinflinty unscrupulous owners and managers who don’t care about the consequences. Keep researching but don’t tout it as skilled helpers which it is not.
You only have to work a little bit with AI to figure out it is a poor research assistant, and training it doesn’t help. Nothing it produces can be taken on trust. It must all be ruthlessly fact checked…so one could argue that you may as well do your own research….
Cam.
I’m sorry, I meant to keep researching AI to make it reliable and safe before releasing it to the public. Doing your own research goes without saying even though I did.
For me it’s “No” to all 3 questions.
I heartily agree with your assessment above. The benefit from AI research is quickly eliminated by the inability to trust it at face value and the resource expenditures needed to verify the accuracy of its results. (You would be better off to skip the time/cost of the AI effort to begin with.)
Just as a computer generated painting has no artistic value to me, an AI generated book of fiction lacks the human author’s unique imagination that I value in a good story.
Machine Learning (Do we really have true AI in the public’s hands at this point?) is a valuable tool for appropriate applications. That category of tools should be constantly improved.
“Thinking Machine” research should be kept in the labs and not prematurely rolled out in cell phones and search engines.
Unfortunately, the corporations see AI as a way to get rid of expensive human labour, which is why it’s being thrust upon us, these days. I’ve switched off Copilot, which I was “gifted” with Windows 11. At least, so far, we’re given a choice.
But the machine intelligence is already embedded in many things, including my spell checker.
Cam.
I’m not interested in a written by, generated by, or organized by A.I. and never will be. I do like my personal router for my own personal organization, spell checking, storage, and writing..