WotC Would you buy WotC products produced or enhanced with AI?

[The Firebird]

When I ask google for information on the grave cleric, sure, I get some links that break copyright laws by printing the entire text. But I mostly get people's opinions on best ways to build a grave cleric and how it compares to other domains. I'd have to do a very specialized search to actually find a copy of Xannathar's to download, though. In other words, I can't illegally download this book without going out of my way to do it.

I don't think this is true, unless adding 'pdf' to the end of the name is very specialized. Even "Xanthar's Guide to Everything" gives me the full text as the second result, after D&DBeyond.

When I asked Gemini "tell me about the D&D5e cleric" and "tell me about the D&D5e grave domain," I got what was pretty much the book's actual text, with very minor alterations that didn't alter the mechanics, even though Xannie's and the grave cleric aren't in the OGL. And all I had to was ask it to tell me about the class and domain.

So you can try to defend it all you want, but Gemini violated the OGL. The fact that other webpages do the same thing doesn't make it OK.

Do you think that Gemini had access to that material because it has been trained on copywritten data, or is it sufficient to have access to the internet? That is, if it never actually saw the Xanathar's book but was able to trawl Enworld, Reddit, the D&DBeyond forums, and anywhere else, would it be able to reconstruct those mechanics?

 

[The Firebird]

The part I quoted.

As to what part is a problem?

Upload of material that one does not have the right to do.
Indexing of said material that is not open/free material.
Serving up hits and links to said material that is not open/free.
The implication that 'its done so whatever' when people are being deprived of compensation for their efforts.

Do I weep over Hasbro materials getting uploaded? Not particularly, as I doubt its really hitting their bottom line in any real way, because frankly the current model of D&D isnt the big winner anyway, its MTG.

Do I feel bad for solo creators who have had their work distributed for free, costing them actual money? Yep.

I don't follow. The last thing you quoted was "Does any of that seem wrong".

I definitely get your point that indexing, serving up links, etc. about that material is wrong.

How do you feel about discussing copywritten material in a way that identifies it? E.g., discussing mechanics in a forum post. For example, if someone said "well, the Grave Cleric gets Bane and False Life at 1st, and Gentle Repose and Spiritual Weapon at 3rd". Bane is a great spell for...you'll want to use Gentle repose to...

Reading enough of these, it would be straightforward to reconstruct the original mechanics. Ok or no?

 

[mamba]

Is this more information than you'd be able to find with a quick google search? I suspect google can direct you to pirated copies of the book and profit off the ads in the meantime.

so because I can find pirated copies it now is ok that AI tells me pirated information?

 

[The Firebird]

so because I can find pirated copies it now is ok that AI tells me pirated information?

People keep interpreting this into what I say. I've not claimed this.

 

[mamba]

People keep interpreting this into what I say. I've not claimed this.

given your statements, it seems like a logical conclusion to me

 

[Scribe]

I don't follow. The last thing you quoted was "Does any of that seem wrong".

I definitely get your point that indexing, serving up links, etc. about that material is wrong.

How do you feel about discussing copywritten material in a way that identifies it? E.g., discussing mechanics in a forum post. For example, if someone said "well, the Grave Cleric gets Bane and False Life at 1st, and Gentle Repose and Spiritual Weapon at 3rd". Bane is a great spell for...you'll want to use Gentle repose to...

Reading enough of these, it would be straightforward to reconstruct the original mechanics. Ok or no?

Much of that would fall under fair use I would imagine, but I'm not a lawyer.

One of the benefits of fair use, is that it can then give enough information, that someone can make a more informed choice to purchase a product.

The whole ecosystem of youtube influencers, essentially being free marketing for Wizards for example.

I would hope you understand the difference between that, and searching up a PDF, or even getting an AI to regurgitate said PDF 'in its own words'.

 

[The Firebird]

given your statements, it seems like a logical conclusion to me

I don't think piracy is ok because google et al. profit off of it. I explained my stance on piracy in general earlier on. So I'm not stating anything like "this is ok because google does it". I think this is ok in general.

If I thought it were not ok, I don't see how I could apply that stance to LLMs but not to google...or not to reddit...or even not to enworld, where these kinds of discussions are common.

But I don't have that view, so I'm wondering how people who do draw this distinction.

 

[The Firebird]

Much of that would fall under fair use I would imagine, but I'm not a lawyer.

One of the benefits of fair use, is that it can then give enough information, that someone can make a more informed choice to purchase a product.

The whole ecosystem of youtube influencers, essentially being free marketing for Wizards for example.

I would hope you understand the difference between that, and searching up a PDF, or even getting an AI to regurgitate said PDF 'in its own words'.

I'm not asking about the legality. I'm asking what you think about the ethics.

Suppose I have a player who wants to play Grave Cleric, but I don't have the book. I look through the public, fair use reddit posts about Grave cleric, and pull out what people say so that I have the mechanics. Then I give them to the player.

I don't think this is a weird example. I've done things like it before; if I forget a mechanic and don't have the book, its typically in online discussions.

 

[Scribe]

I'm not asking about the legality. I'm asking what you think about the ethics.

Suppose I have a player who wants to play Grave Cleric, but I don't have the book. I look through the public, fair use reddit posts about Grave cleric, and pull out what people say so that I have the mechanics. Then I give them to the player.

I don't think this is a weird example. I've done things like it before; if I forget a mechanic and don't have the book, its typically in online discussions.

At a certain point, it becomes "I know it when I see it."

There is no equivalency between what you have described, and what these LLMs are doing.

The scale is so fundamentally different, its scratching numbers in the dirt vs quantum calculations.

If you are able to divine out of a bunch of discussion threads and posts, the rough form of a Grave Cleric, am I going to say its the same thing as ripping a bunch of text out of PDFs and blending it up in an LLM?

No.

 

[Faolyn]

I don't think this is true, unless adding 'pdf' to the end of the name is very specialized. Even "Xanthar's Guide to Everything" gives me the full text as the second result, after D&DBeyond.

Yes, adding pdf is specialized, because 5e books are not typically sold as pdfs. The only way to obtain a 5e pdf is through scanning it in. And sure, if you just google "Xanathar's Guide to Everything," you get a link to D&D beyond... where you can buy the book for $29.95 (digital only--not a pdf). If you google "Xanathar's Guide to Everything pdf" you get the Internet Archive, Anyflip, Scribd, and other sites where you obtain a pdf of the book without paying for it.

But when I type in "savage worlds pdf," I get a link to the actual PEG website, because those books are also sold in pdf form.

As Scribe said, I'm not weeping if people do this with Xannie's because Hasbro can afford to eat the loss. But that doesn't mean it's right.

Do you think that Gemini had access to that material because it has been trained on copywritten data, or is it sufficient to have access to the internet? That is, if it never actually saw the Xanathar's book but was able to trawl Enworld, Reddit, the D&DBeyond forums, and anywhere else, would it be able to reconstruct those mechanics?

We know that Meta, at least, was trained on copyrighted data. We have actual proof of this. There have been lawsuits filed. Apparently (as of two days ago) Meta is claiming that the 7 million books it pirated had "no economic value" and that they're protected under "fair use" because, they claim, they don't reproduce the entire book.

Now, I got Gemini to pretty much reproduce the entirety of the grave cleric, which is not OGL. Which means that the idea that AI won't reproduce copyrighten material is bogus. Maybe some AIs won't, but others will.

So we have two options here:

(1) Gemini, like Meta, was trained on stolen material. This is morally, ethically, and financially wrong, no matter what (generic) you think about WotC or Hasbro.

(2) Gemini wasn't specifically trained on stolen data, but got its info from the internet. In this case, Gemini is dangerous because it's grabbing things randomly from online, which means its going to be giving false or even harmful information... such as in the case of the poisonous mushrooms. (I admit I don't know which AI wrote that book.) While a D&D character isn't going to kill someone, it could--since it's picking information randonly instead being trained--give the wrong class information, which could lead to in-game problems.

And unless you have some evidence that Meta, ChatGPT, or whatever other AI you're using is better trained, smarter, or whatever, I see no reason not to assume that that other AI programs are trained in the same way.