and they are also very likely to receive the bunko treatment at one of the next summer or winter fairs.
Sometimes the best way to get your level up is by diving into hard stuff you are interested in. Whenever they turn into paperbacks I wish you the best of reading luck.
Just like omk, I will wait for some time until they become somewhat cheaper.
plus, there are so many previous winners that have become cheap already
So apparently the 東京都同情塔 author openly admitted to using AI while writing her novel.
Relevant question in the press conference at the 5:20 mark:
By the way, since it deals with AI and crime, I wonder to what extent it’s similar to ユートロニカのこちら側 | L38 (learnnatively.com)…
Oh-ho-ho. Will need to give this a watch when I have some time.
She shouldn’t be allowed to win with something she used AI to write
Also I am avoiding all her works like the plague now
This was my first reaction as well, but the book is about AI in the first place, so it’s not strange she used actual AI as inspiration. She says that about 5% of the book is straight from AI, which sounds like an awful lot, unless it’s strictly for things AI-related in the book itself.
In another comment she sounded to me like she used AI as a sounding board, a thing to bounce ideas off and see how she finds them. Hopefully it’s not as bad as it initially sounds.
I’d be interested to look into this further. Still, a strange thing to admit openly, and sure to cause controversy.
I get that it’s experimental, but like @enbyboiwonder it rubs me incredibly the wrong way. ‘Just because you can doesn’t mean you should’. I want meaning behind the words I read for literature, not computer word vomit.
Using AI as a sounding board to me is whatever, people use all sorts of crazy things as means to get their thoughts unstuck, but having straight AI output being entered into the book for me to read is a hard pass.
Unless it’s dialogue with a chatbot in the book too?
That’s the only excuse I can find for using text straight from a chatbot.
Seconding all your sentiments here. My initial reaction was “No way am I going to read that book!”
Especially wondering how they even considered giving her a prize for that…
I briefly thought so too, but that doesn’t match the “bouncing ideas” bit in my mind
Yeah, that’s the only passable excuse I can think of too, but unless she explicitly states that (I haven’t watched the interview) I kind of doubt that’s what is happening. It’s also just strikes me a bit…uninteresting? Like I’ve seen plenty of dialogues between humans and computers online, I don’t really have a need for that in books I read unless the author has made the conversation special somehow.
I suspect the controversy is the point – probably more publicity and more sales than if she’d just done a standard winner’s speech.
With the risk of discrediting her work though?
Especially since she said she’ll do it again…
My first completed book of 2024 was 火花 | L35 and it must be a sign that my year in reading is off to a great start because I absolutely loved it and awarded it with a coveted 5 stars! Even though this is one of the most popular Akutagawa winners from recent years, I just never got around to reading it until now. I don’t really find 漫才 particularly funny and think a lot of it gets lost in cultural translation, and since it is the topic of the book it kind of put me off. I turned out to be very wrong, though, since this book was about more than just 漫才. In fact, I may have a new appreciation for it now.
I think 火花 is the type of book that not everyone will love, but I found the character growth, philosophical discussions on what it means to express yourself, what is funny or not, or how to make a living as a comedian and artist, etc. fantastic, and the fact the author himself is is a comedian gave the writing dimension and made the homage at the end of the story all the more beautiful.
Anyways, I will stop raving here and leave my review (as I am trying to write more reviews for what I read in 2024).
Wow, kind of surprised by the vehemence against an author using AI generated words in their book… if you think about it in terms of sentences, 5% is only one out of every 20 sentences. And presumably the author chose what to keep and what to scrap, so I would never assume that the book therefore automatically lacks meaning behind the words.
There are lots of experimental writers using lots of different techniques to create randomness or interest… A book recently came out that was made out of years of the authors journals, transcribed and then each sentence sorted alphabetically. “Found” poetry created out of scraps from newspapers is a thing.
I don’t see how AI generated text (used in moderation) is substantially different from these other forms of experimentation.
To me I view in the larger context of AI being used to replace jobs, particularly creative jobs, as well as a loss of the emotional art connection. If other people want to read it I won’t condemn them, but I have no interest in engaging with it as art.
I further find it deceptive to not release this info until immediately after winning the prize.
This article has more information about how Rie Kudan specifically used chatgpt in writing 東京同情塔, and I thought it was an interesting read: AIが”生んだ”?芥川賞「東京都同情塔」 作家・九段理江さんが明かした誕生秘話 | NHK | WEB特集
She mentions that the character in the book chats with an AI chatbot similar to chatgpt, and she used chatgpt to generate the actual responses. A big theme of the book is also how words are used and change meanings, so she asked chatgpt for alternatives to the word 刑務所. She also used chatgpt to edit the text for consistency, and more. In short, it seems more like she utilized it as a tool (while directly utilizing the output in cases where it edited her text) and for creating authentic responses from the chatbot in the book.
Considering the book is about AI, I think the fact chatgpt was used like that adds something to it as a work of literature in this very specific case. However, I don’t know if that would be the case for the vast majority of books that aren’t about AI. I’m neutral on this topic still and waiting to see how the literary world develops from here on out, but I am curious about people’s thoughts on using chatgpt as a brainstorming or editing tool and not for writing the entire book? If chatgpt edits something, is it ok to take that text? When does it stop being your own creation?
@omk3 actually floated the idea of would it be OK if it was actual chatbot conversation in the book and I replied it was OK if boring.
I also have no issue with people using AI to bounce ideas around to get their creative wheels turning (think that was also upthread?).
With editing its a bit less clear cut. Editing like how Google docs will suggest a different word choice or editing as in a complete rephrasing of a paragraph? The former feels akin to using a thesaurus, the latter I’m less keen on, especially if the changes are swallowed whole.