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

[Scribe]

You know what you could use AI for: to write SRDS for AD&D and 4E...

Ask ChatGPT what Thac0 is in natural language.

 

[mamba]

So like a person who reads a bunch of stories and then is inspired to write their own story

no, not like a person, it is not inspired, it regurgitates based on heuristics

 

[The Firebird]

no, not like a person, it is not inspired, it regurgitates based on heuristics

It doesn't have to work as a human does for the principle to apply. If you take hundreds of works and process them into something new, then the new thing is something original.

 

[mamba]

It doesn't have to work as a human does for the principle to apply. If you take hundreds of works and process them into something new, then the new thing is something original.

I don't think it is original enough to clear the hurdle you set up earlier

Yeah. And I watched Star Wars.

You know, I don't think Lucas lists Kurosawa or Leone in the credits of Star Wars.

 

[Incenjucar]

If it's still there but extremely difficult to untangle and most probably incomplete given the loss of information, would you really judge a tool on what it can do under an extremely difficult scenario to implement? I can probably do all sort of illegal actions with a spoon, if I try enough, yet spoons are OK where I live, because their normal use case makes them sufficently useful to keep allowed, despite the inherent risk of having everyone able to misuse a spoon. One might have the opinion that the benefits of the tool don't outweigh the cost, or the other way round, but I don't feel that banning a tool unless it can't have a harmful usecase is generally a good idea. We'd be using very few tools.

You're asking someone in cybersecurity to shrug off using technology for evil.

My retirement is guaranteed to be cozy because of the ways GenAI can be used to do harm.

 

[Jfdlsjfd]

You're asking someone in cybersecurity to shrug off using technology for evil.

My retirement is guaranteed to be cozy because of the ways GenAI can be used to do harm.

I was asking you whether you think it's worth banning computers or the Internet because they can be used for cybercrime, to be exact. I disagree with your view that allowing a technology means "shrugging off" using it for evil. It's not because cars are allowed that you should necessarily shrug off people being run over (though it is certainly a way to deal with the problem in a low-regulation society).

 

[Incenjucar]

I was asking you whether you think it's worth banning computers or the Internet because they can be used for cybercrime, to be exact. I disagree with your view that allowing a technology means "shrugging off" using it for evil. It's not because cars are allowed that you should necessarily shrug off people being run over (though it is certainly a way to deal with the problem in a low-regulation society).

If you misuse computers you can have your right to use the internet taken away and be thrown in jail etc. etc. You're asking a really bad whataboutism.

Making RPGs easier to run for a subset of the gaming population is also not comparable to things like curing cancer or powering global economies or even allowing households to do their taxes without basic math errors bringing the IRS down on them. It is best to keep this stuff in the context of the discussion.

The models used by content generators are largely obtained in an ethically dubious-at-best way. Its value as a creative resource is pretty mid, but there are applications if the ethics don't bother you (or, ideally, if a properly-licensed model is created). I personally still wouldn't use it even if the ethical concerns were resolved because I have developed my artistic skills and am more interested in human art than remixed content.

 

[Lanefan]

Thank you. It's nice to see that someone else here has actually researched this.

We need better STEM education.

What's STEM, in this context?

 

[Sorcerers Apprentice]

AI models do not contain the complete text of the works they have been trained on. That would be physically impossible given that commercial Large Language Models are trained on trillions of words yet only require as much disk space as billions of words. These models encode the concepts, the relationships between words in a text, not the text itself. Only some famous quotations appear often enough in the data set to be fully recorded.

 

[Jfdlsjfd]

If you misuse computers you can have your right to use the internet taken away and be thrown in jail etc. etc. You're asking a really bad whataboutism.

Sure, like any activity. What make you think that using AI for bad thing can't lead one to go to jail? Allowing the use of a tool doesn't mean that it should be totally unregulated. Many tools are allowed in almost all countries of the world, and yet are subject to regulations.

Also, You not understanding my analogy to illustrate my point doesn't make my point a whataboutism. Let me rephrase my point for easier readability:

When considering the use of tool X, irrespective of the value of X, an instance of X being used to harm isn't enough to justify the banning of X. The banning of X should happens only when the drawback outweigh the overall benefit of the let X being freely used. Also, as a side note, there is no reason to expect that all lawmaking bodies will agree on the calculation on the benefits and rule in the same way as a result.

Whenever we determine that outcome Y is harmful, it is much more efficient for the lawmaking body to pass a law to punish people causing outcome Y, rather than passing laws banning tool X (and Z, and N...) because these tools can, among other uses, be used to cause Y.

This is true for any tool X and outcome Y. Saying that X can take several values (cars, guns, medicine, photocopiers, the Internet, clothes, coffee machines, lego bricks, basically anything since that's the point) to illlustrate the point doesn't make this point a whataboutism, it's using an example to show that very dangerous technologies can be allowed (even if they are regulated, as obviously "allowed" doesn't mean "escaping all regulation") despite constant evidence of the harm they do, to show that indeed, pointing out a case where a technology can do harm hasn't resulted in it being banned. If any mentions of cars was confusing, I am sorry about it and I retract them.

Making RPGs easier to run for a subset of the gaming population is also not comparable to things like curing cancer or powering global economies or even allowing households to do their taxes without basic math errors bringing the IRS down on them. It is best to keep this stuff in the context of the discussion.

On the other hand, if it is forbidden as a technology, based on the fact that it could allow WotC to produce cheap RPG modules, I fear it may also hamper the development of other uses. The legislator must take everything into account when determining if regulation is needed, and one of the motives given in the act for allowing a text and data mining exception to copyright by the EU lawmakers was to create an economic environment that is favourable to AI development in general, not because they wanted people to be able to draw pretty pictures on their character sheets. I don't think having a law saying "AI is allowed everywhere, except that using it for the context of RPG gaming is punished by 5 years of jail and a 100,000 € fine" will fly, though that's certainly something one can advocate if that's the only problem the majority of the people find with AI, and every other use case is positive.

I personally still wouldn't use it even if the ethical concerns were resolved because I have developed my artistic skills and am more interested in human art than remixed content.

Sure, more power to you! I have absolutely no qualm with people not liking AI products. "I don't like it" is perfectly fine and can't really be questionned, "it should be legally forbidden" is another level of demonstration.