Now you can disclose whether your WIP uses AI

I’m anything but an AI advocate…but I don’t think this particular criticism holds up unless you’re already refraining from music and video streaming, Zoom calls, and cloud computing, all of which consume significantly more energy and water than an AI query.

I think it’s a great argument for refraining from using Bitcoin, but as I noted on the other thread,

It’s also not clear to me that Google Translate or Grammarly’s spellcheck were trained by unfair use of copyrighted material. Use of a machine learning algorithm isn’t what I object to about GenAI. Institutionalized large-scale IP theft is. I don’t think we’ll solve that by a Butlerian Jihad.

Edit: and with OpenAI losing a chunk of money for every non-subscriber who uses ChatGPT, it’s not clear to me that an anti-AI campaign should be calling for a boycott on queries, rather than hastening the day when OpenAI runs out of money and has to pull back most of its AI services to be subscriber-only.

18 Likes

But where do we draw the line then? I was chastised for asking a chatbot questions earlier in the thread and translating a few paragraphs to French (and I’m fine with that, I understand it was wrong now, I just didn’t know better). But Grammarly, which is AI too is okay as long as its just used a little? I’m really trying to toe the line and avoid AI, but I need a boundary here.

I’m not sure what boundary you’re asking for that isn’t covered by this post - I’m not trying to be snarky, just hoping to be able to be as clear as I can if I haven’t been earlier.

(If you’re asking about ethical line-drawing, that’s a separate matter and is maybe better suited to the general AI discussion thread.)

3 Likes

I’m referring to the post saying to only use Grammarly as a spell checker. The spell checker is AI too, but that’s apparently okay? I’m sorry, I’m so confused.

And I should have spoken up to defend you then – sorry. But the people chastising you weren’t the ones who draw the lines for COG. They’re a chunk of the audience with strongly-held views, and you should gauge their feedback accordingly.

I understand how interacting with the strongest views on this might lead one to

but I’m certainly not going to do that, and I don’t think anyone else should be forced to.

I agree with Harris that the strongest reason to avoid ChatGPT for research is its unreliability. I’ve yet to see AI research output in areas I know anything about that didn’t include significant inaccuracies, even if not outright hallucinations.

ChatGPT being trained on the moral equivalent of theft is another reason to refrain… but as I noted above, while I think that’s more than sufficient reason not to pay for it, I suspect that non-subscriber use at this point at this point hurts OpenAI more than it helps them. So the best action to take on that basis is debatable.

There is plenty of non-copyrighted, open-source content out there for purposes of AI training, and my understanding is that it was adequate for loads of machine learning functions, including Google Translate. LLMs have ended up demanding more writing than has ever been produced in all of human history, and a bunch of companies have responded to that by pushing “fair use” past what ought to be a legal breaking point, if not ignoring it entirely. But not all AI is based on LLM-style training.

CoG’s caution about Google Translate is based on the inaccuracy of its translations; the rule isn’t “don’t use it,” it’s “don’t rely on its output without checking with someone fluent in English.” They note that it works OK for short articles, which is not what you say if you’re ruling it out entirely on moral/legal grounds.

Their caution about Grammarly isn’t about the spell/grammar check, which they call “fine,” but about using its specifically generative functions. That can lead to the AI (re)writing large chunks of text for you, with the ensuing problems around “sounding like an AI” and not being copyrightable which underlie the CoG/HG policy of not publishing text generated by AI.

With Harris, I think that line’s pretty clear. There’s a vocal share of the CoG/HG audience who would like the line to be drawn at “you must never have used any AI for anything, ever,” but that’s their line, not CoG’s, and should be treated like the opinion of any other chunk of the reader base with a strong opinion.

24 Likes

So can we use spellcheckers at all now? Or is that also included? I have never touched generative AI (I hate it with a passion), I have gone to all possible lengths to use programs that do not include it (old Word, for example, not the new shitty one) but I do use spell and grammar checkers (mainly Grammarly, you can turn off a lot of suggestions) to flag errors I can fix (no wholesale rewriting of sentences, why the hell would I want that).

I guess my question is: If I would have continued to use the new Word instead of telling Microsoft to roll me back to the old one, would I be an AI user? If I use Grammarly for spellchecking, am I an AI user? In both cases, the AI functionality is something that can be turned off in the program, so I never need to see or use it. And even worse… with how much the tech word pushes AI functions on us, more and more programs have the AI functionality in them.

In other words, is it the USE of generative AI that needs to be disclosed? Or using a product that has potential AI functionality that needs to be disclosed?

Sorry for being a bit anal about this, I hate AI with a burning passion, but I am trying to understand what is actually being asked of people here. Because I can stop spellchecking, not sure y’all would appreciate it though…

17 Likes

I’d usually hesitate to answer a policy question when I don’t in any way speak for the company, but it honestly seems pretty clear to me that the answer is:

In CoG’s eyes, no. (Try to read these policies as if they’re ruling out use of spellcheckers or Grammarly or Word. It just doesn’t make sense.)

In the eyes of a share of the audience who also hate AI with a passion, yes.

Do with that what you will.

2 Likes

Yeah, I am just slightly more autistic than usual today I suppose. :grin: Now that I look again if clearly says that it is text GENERATED with generative AI, which I won’t touch with a stick. I can manhandle the english language fine on my own thank you very much. And it’s not like my writing style sticks to the rules very much, I just like Grammarly because it can help me flag some stuff I usually mess up so I can decide whether it was a mistake I want to fix, or a stylistic choice I want to use.

EDIT: To be fair I rarely spell or grammarcheck my demos either… :melting_face: that’s for later.

EDIT: Even more annoying is that so many programs now label themselves as AI-powered as a selling point, while not being generative AI. AI as a term now has become so vague it is nearly unusable. While Generative AI and LLM’s are a very specific thing…

EDIT: And for anybody using Grammarly (not sure if this is just in Grammarly Pro), I would recommend to go in and check the settings. You can turn off all the suggestion stuff, while keeping the checking stuff. For example:


12 Likes

Thanks @Havenstone, I appreciate it. I googled Grammarly’s spellchecker though and some sources said it does use AI - specifically the spellchecker too. Hence my confusion and me irritating everyone on this thread with my repetitive questions.

1 Like

Just to say I’m not irritated, just wanting to help with clarity. I hope Havenstone and Malin’s additional discussion has been helpful!

6 Likes

Thank you. I’m going to be painfully honest, while I’d love to use Grammarly because my story could benefit from a good spellcheck, I’m too scared to. I’m afraid that anything remotely connected to AI can now make my story a target for a witch hunt.

But I really do appreciate your clarifications and everyone trying to help me.

3 Likes

But not generative AI. I’ll be very impressed if you find a way to generate a narrative text using just a spellchecker.

Meanwhile, I hope seeing a would-be author intimidated out of using a spellchecker helps some people reconsider the breadth of the brush with which they’re tarring machine-learning software products.

24 Likes

I’ve had (pre-AI) Word’s spellchecker’s suggestions sometimes being so wildly inaccurate that using them would change the text’s meaning completely, which isn’t that far off. But that’s mostly Word’s Finnish vocabulary being nonsense.

3 Likes

The autocomplete on my phone regularly assumes I meant to type “bogs” instead of “boys,” “frog” instead of “from,” and other similar ridiculousness where it picks the obscurer or obviously less-correct-in-context word. Maybe it’s just been trying really hard to turn my SMSes into a novel about swamp denizens and I should just let it generate.

8 Likes

23 Likes

…I suddenly want to write a horror story about Sleeper in the Green who has infiltrated electronics to get summoned by careless typing.

2 Likes

Thanks for the tips! I’d like to share what I’m doing to comply with this new AI disclosure. I previously used AI to standardize the spelling into UK spelling (e.g. humor → humour, realize → realise, and so on) This meant that I currently would need to tick the ‘yes’ box. Luckily, I had these edits done in three separate updates, so it was easy to revert all of them.

For the next three to four weeks, I’ll be going through my entire text by hand and reading through it for US spellings, and changing them manually to UK spelling. This way I won’t have included any output of generative AI in my game. I don’t mind doing this — it’s a good opportunity to read through it and look for other errors — but if I knew this would be a thing, I could have avoided this completely. I hope Choice of Games can keep writers in mind as they continue to evolve their policies around AI.

1 Like

Fortunately, you can do that with find-and-replace.

7 Likes

Pretty sure this isn’t aimed at spell checkers or programs that suggest the odd comma or change in word for clarity (which if you use word is usually wrong anyhow and can be safely ignored). It’s more programs that are rewriting entire sentences/paragraphs.

^This. “AI” is the new trendy thing. They’re using it for plain old normal computing programs that have often been around for ages without the AI label in the description because it now sounds “cool”. Generative AI is a whole different kettle of fish.

6 Likes

I’m not sure this kind of clearly excessive response is needed or expected. I think what’s happening is that you are mixing up CoG’s position with that of a section of this forum’s userbase.

If I understand correctly from the recent discussion here, the new policy does not say that you “need to tick the ‘yes’ box” if you use software to automatically standardize spelling between different localizations. If you just go back and re-read the new policy again, CoG’s boundary is clearly at using generative AI to produce content for your story, not at using spellcheckers like Grammarly or Word’s built-in functionality. The line at using any form of spellchecker or translation software in any capacity is not from CoG, it’s from users of this forum.

But if you really want to do that, a simple find-and-replace will work great, there’s no need to force unnecessary toil on yourself.

6 Likes