D&D General Would you buy an AI-generated Castle Greyhawk "by" Gary Gygax?" Should you?

[Mecheon]

2) I've seen the Terminator movies and Black Mirror. The day will come when the machines will care about who has said "please" and "thank you" to them.

Yes, I also have heard of Roko's Basilisk but find it rather dumb as it really goes hard oon the 'oh the AI will just do this clearly" when I frankly think the main thing that'll happen is the AI being shut down the moment it freely judges efficiency and tries to fire a CEO

We have an VR AI hell these days, its called "VRChat rollercoaster and its effects on the human inner ear"

I find the arguments about the quality of AI work being the reason for not buying it unconvincing. If folks are buying it, then obviously the quality was good enough for their purposes, and who am I to say they are wrong?

People spent money on NFTs, so, yeah, quality won't stop someone. But, the thing is, we have a Castle Greyhawk that exists already. Its called Castle Zagyg. You can just, buy that and convert to D&D

Given we have the 'Actually written with Gygax's input', we can compare. And AI will be found wanting

There is an argument that generative AI costs jobs. This is true. Automation costing people jobs is not new, though, so really this comes down to our fears that a new kind of job is at risk. These robots aren't coming for the assembly line jobs (which had already come for the cottage industry jobs), now they are coming for the office jobs, and the art studio jobs.

And you know what's been happening? Its producing lazy and false things. For example, one website decided to just use AI to write articles rather than hire people. And guess what? They were false articles. Blatently so. And folks found the source. So, folks do what folks do best and fed the machine with more falsehoods, which is how we got the Glorbo craze. We already have enough issues with SEO pushers making search engines useless due to flooding everything with garbage, AI is just a tool for them to produce more garbage. Speaking of, here's artwork done of a recently discovered early croco-relative, something never produced in art before. And here's the garbage some lazy museum spat out with AI that, you might notice, doesn't look anything like what it looked like. Almost because.... Using AI rather than artists is lazy and shows you don't care about the truth. Or facts. Or you want to confuse real actual scimitars with their D&D counterparts so it just scrapes data from D&D when talking about the real, actual weapon.

We're allowed to, y'know, fight back against this wave of blatent and stupid misinformation that's gotten people in serious legal trouble when they decided to use ChatGPT to do legal stuff for 'em. Just sitting back and going 'oh its the future' at such comical, ridiculous mishaps like this is just, letting this get through. Is that what people want? Blatent falsehoods and more information garbage stinking up the internet?

We embarassed a website so hard it deleted every single WoW article it ever 'wrote' with the Glorbo debacle. If folks keep pressure on people using AI, if they're reminded every time that people can recognise it, that its lazy, that it presents a bad image, all things that are absolutely 100% true at the moment, then it won't be the future, it'll just be a tool that's best used for what AI's good at, data analysis.

EDIT: Or maybe AI will just kill people.

What we need to be asking is not, "is AI art good or bad quality?" since that is both subjective and rapidly changing. Nor is it worth arguing about whether or not it is going to happen or should happen. It has happened and isn't going to go away.

What we need to be exploring is what we can do with it to make our world better, and how we can ameliorate the harms that inevitably ensue with rapid change.

AI isn't making the world better. Its mass-generating garbage masquerading in the shape of truth. We should be dunking on these falsehoods and reminding people they're not getting away with it

We have a Castle Greyhawk already, its Castle Zagyg. An AI one is just going to be what happens if you throw Gygax's other modules in a blender and will fail to live up to anything people want

 

[QuentinGeorge]

Yes, I also have heard of Roko's Basilisk but find it rather dumb as it really goes hard oon the 'oh the AI will just do this clearly" when I frankly think the main thing that'll happen is the AI being shut down the moment it freely judges efficiency and tries to fire a CEO

We have an VR AI hell these days, its called "VRChat rollercoaster and its effects on the human inner ear"


People spent money on NFTs, so, yeah, quality won't stop someone. But, the thing is, we have a Castle Greyhawk that exists already. Its called Castle Zagyg. You can just, buy that and convert to D&D

Given we have the 'Actually written with Gygax's input', we can compare. And AI will be found wanting


And you know what's been happening? Its producing lazy and false things. For example, one website decided to just use AI to write articles rather than hire people. And guess what? They were false articles. Blatently so. And folks found the source. So, folks do what folks do best and fed the machine with more falsehoods, which is how we got the Glorbo craze. We already have enough issues with SEO pushers making search engines useless due to flooding everything with garbage, AI is just a tool for them to produce more garbage. Speaking of, here's artwork done of a recently discovered early croco-relative, something never produced in art before. And here's the garbage some lazy museum spat out with AI that, you might notice, doesn't look anything like what it looked like. Almost because.... Using AI rather than artists is lazy and shows you don't care about the truth. Or facts. Or you want to confuse real actual scimitars with their D&D counterparts so it just scrapes data from D&D when talking about the real, actual weapon.

We're allowed to, y'know, fight back against this wave of blatent and stupid misinformation that's gotten people in serious legal trouble when they decided to use ChatGPT to do legal stuff for 'em. Just sitting back and going 'oh its the future' at such comical, ridiculous mishaps like this is just, letting this get through. Is that what people want? Blatent falsehoods and more information garbage stinking up the internet?

We embarassed a website so hard it deleted every single WoW article it ever 'wrote' with the Glorbo debacle. If folks keep pressure on people using AI, if they're reminded every time that people can recognise it, that its lazy, that it presents a bad image, all things that are absolutely 100% true at the moment, then it won't be the future, it'll just be a tool that's best used for what AI's good at, data analysis.

EDIT: Or maybe AI will just kill people.


AI isn't making the world better. Its mass-generating garbage masquerading in the shape of truth. We should be dunking on these falsehoods and reminding people they're not getting away with it

We have a Castle Greyhawk already, its Castle Zagyg. An AI one is just going to be what happens if you throw Gygax's other modules in a blender and will fail to live up to anything people want

Most of what you're saying is fine but just one point is that we don't actually have a Castle Zagyg - because Troll Lords never got to actually publishing anything about the actual castle before Gary died (except Yggsburgh and the "Upper Works")....

 

[Whizbang Dustyboots]

What we need to be exploring is what we can do with it to make our world better, and how we can ameliorate the harms that inevitably ensue with rapid change.

Given the importance of written communication, and the fact that ability to write well isn't evenly distributed, I like the democratization we're already seeing, with AI writing tools being added to existing programs like email and other writing programs. People with good ideas shouldn't be disadvantaged because they didn't go to a school where they learned to speak and write the dominant language in their area as well as someone of a higher socio-economic background.

People should succeed based on their hard work and their ideas, not how well they can polish a turd. Generative AI is already helping there.

 

[Whizbang Dustyboots]

Yes, I also have heard of Roko's Basilisk but find it rather dumb as it really goes hard oon the 'oh the AI will just do this clearly" when I frankly think the main thing that'll happen is the AI being shut down the moment it freely judges efficiency and tries to fire a CEO

The CEO has always been subordinate to boards of directors and especially shareholders. If they find out that the CEO is thwarting "business optimization tools" and theoretically costing shareholders value, the CEO who pulls this will be out on their ass.

The AIs are definitely coming for management and very few people will shed a lot of tears.

People spent money on NFTs, so, yeah, quality won't stop someone. But, the thing is, we have a Castle Greyhawk that exists already. Its called Castle Zagyg. You can just, buy that and convert to D&D

Castle Zagyg is a great example, because it's incomplete. A lot of it is just notes. So, would a generative AI -- which we've already demonstrated in this thread can do Gygax's written style -- trained on Gygax's works (or a chosen subset) working off those notes to complete Castle Zagyg be creating something of worth? Would it be better or worse than a human doing it? Would it matter if the AI doing it could do it cheaper (I have no idea how much it would cost the Troll Lords to get an AI Gygax built for them) and could get it to market faster? (Fans of 1E AD&D aren't getting any younger.)

And you know what's been happening? Its producing lazy and false things.

Yep, generative AI isn't ready for many of the tasks it's been given. (At least we're not talking about Tesla's "self-driving" features here.)

But we're still in the very early stages. I'm sure you remember when DVD players were enormous and cost $1,000. Today, if you can find them, they cost $20 and aren't much bigger than a Blu-Ray disc.

In the very near future -- AI tech is improving far faster than DVD tech did -- generative AI hallucinations will be a thing of the past. Even free AI art generators now typically get the number of fingers right in pictures of people, something it famously struggled with a year ago. (The very AI over-saturated colors are still very much a thing, though, even when there's no other obvious tells that art may have been AI generated.)

AI isn't making the world better.

It is already finding brain tumors months earlier than a human radiologist can. That's definitely making the world better.

Lumping all AI uses into what Sports Illustrated is shortsightedly doing is pretty silly.

We have a Castle Greyhawk already, its Castle Zagyg.

Castle Zagyg was never finished before EGG's death.

 

[Scribe]

Given the importance of written communication, and the fact that ability to write well isn't evenly distributed, I like the democratization we're already seeing, with AI writing tools being added to existing programs like email and other writing programs. People with good ideas shouldn't be disadvantaged because they didn't go to a school where they learned to speak and write the dominant language in their area as well as someone of a higher socio-economic background.

People should succeed based on their hard work and their ideas, not how well they can polish a turd. Generative AI is already helping there.

... unless we are talking about a 2nd or 3rd language (Europeans and their better education systems!) what?

You dont need a post secondary education to write emails, and we certainly dont need the idiotic prompts of Outlook/Teams to convey a message.

 

[Whizbang Dustyboots]

... unless we are talking about a 2nd or 3rd language (Europeans and their better education systems!) what?

You dont need a post secondary education to write emails, and we certainly dont need the idiotic prompts of Outlook/Teams to convey a message.

In my workplace, and the workplaces I've worked in, people who struggle to express their ideas often don't get those ideas adopted, losing out in favor of people who can communicate well but may have no decent ideas.

I live in a region of the world with a lot of immigrants, who often struggle with their command of the English language. I see this happening first-hand.

And those idiotic prompts are going to get a lot better very soon.

 

[Scribe]

I live in a region of the world with a lot of immigrants, who often struggle with language. I see this happening first-hand.

So we are talking 2nd/3rd/nth language here.

 

[Whizbang Dustyboots]

So we are talking 2nd/3rd/nth language here.

So?

 

[Scribe]

So?

So that was the qualifier I was using in my response to you. Unless we are talking about...

So, sure. As a tool to assist people for whom English is not their native language, I can see how that is of benefit.

I still refuse to press the "Got it, thanks!" button on Teams.

 

[Whizbang Dustyboots]

I rolled my eyes for a long time at Gmail's attempts to guess what I was writing. And then, about a year ago, it got really good at it and I began just tapping tab to let it complete sentences periodically. I still write the majority of my emails -- and I'm a very strong written communicator in English -- but if I'm finding value in these tools, I think it's likely that they're more widely used than anyone outside Google could imagine.