2. Do you also find adventure games that use 'procedural generation' to be really repetitive and gimmicky?

This is just a personal interpretation, but I perceive procedural generation as somewhat of a spiritual predecessor to generative AI. My first encounter with it was in Indiana Jones and his Desktop Adventures demo. It was fun to play at first, but it soon became apparent how random it all is. Basically it is replay value solely for the sake of replay value. Certainly not the bespoke feeling of a hand-crafted game that I ideally desire.

This has always been my mindset as well. I've never enjoyed games which employee procedural generation in order to create their levels. I want a level which has been carefully and purposefully built with clear intention. An generative AI cannot understand what constitutes a truly well designed level, it can only work within a rudimentary framework of what such a level ought to contain.

1. Why do people use AI to make adventure games rather than using public domain and creative commons assets?

Not every kind of asset you need for your game is available as public domain or creative commons asset, when you have a specific style in mind. So if they decide to use AI generated assets instead, it is probably because they think it better matches their vision. And it might be more (time) efficient to work based on (or with) a generated asset than for example taking a public domain asset and paint over that to make it match the planned style.

And for some it also might not be a concern (ethically) at all where the asset has come from and how it has been created. They may not value an asset more when it has been created by a human, it may not be a priority for them to create a truly hand-crafted game.

Before AI, people just stole or copied assets from the internet when they needed something, cropped and used these directly. Using AI might be a lower entry level for stealing, giving them the feeling that "they have created the image by themselves" (while the AI automatically did the stealing for them invisibly in the background).

2. Do you also find adventure games that use 'procedural generation' to be really repetitive and gimmicky?

I didn't know that Indiana Jones desktop adventure included procedurally generated levels. Never played it. I have never seen procedurally generated content in an adventure either. Implementing procedurally generated levels just for stretching playtime is annoying, when the whole game isn't about dungeon crawling. Labyrinths can be really annoying.

But still, hehe: I like the idea of rogue-like games. These are not the kind of "adventures" we are usually talking about here, but for me they are still also adventures. And they fascinate me somehow. It's not like I would play them for hours, but for example in a game like "Pixel Dungeon" I enjoyed them. Or "Diablo", which was inspired by the old classic "rogue". It is just a different kind of game, and for me it is totally fine to have the surprises, sometimes have a good run, and sometimes not.

What kinda fascinates me about the procedural generation is that you need a stable ruleset to prevent the generation to go wrong. To have an entry and exit, to have enough content so it doesn't get boring, to have a "survivable" path in enemy difficulty, if there are enemies spawned ... in Pixel Dungeon, you randomly find items, and while the look is the same, in each round they can have different effects (randomly), and you have to find out first what the items do this time, after trying it out. It is a surprising element each time, and makes the game feel fresh and less boring overall.

And: I am fine when procedural generation is being used for secondary "background fill" that doesn't really matter. Like a ground texture or a wall, or a shelf full of different books. Simply patterns to prevent something looking empty.

For me it is totally fine if a game works like this, but it should also be sold and presented as a game that included such elements. To prevent players with different expectations to be disappopinted.

Had you already read the interview where Revolution Software said they were doing in-between frames with AI technology?
https://www.polygon.com/23842925/broken-...les-cecil/

Have a quick look at the clown picture in this next article on Medium. I can't tell if the clown's teeth are the result of human error or AI-induced, either way it does look quite strange.

Hard to tell. Like a forgotten outline. Maybe a disabled layer ... I think it is something that can happen when humans do it as well. Like a slip. A "comic drawing" teacher of mine once published a comic he has created and realized after it was printed that he gave one character accidentally 6 instead of 5 fingers. Exactly the most "popular" mistake everyone knows from AI. But humans simply make mistakes as well, even when they have been professionally drawing for years.

When someone is distracted and doesn't feel well, someone is angry or very sad, or under horrific stress, out of time, sometimes they cannot work as perfectly as they should. I don't know what's behind the clown face, and i won't speculate against the creator in such a case.

I'd say no to AI usage in AGs because we all know how business works: if a business practice is considered "cost effective" and there is no opposition to its 
implementation - it gets spread to all parts of business.

Before long we'll have AI-generated music, AI voiceovers, AI-written game scripts in addition to AI-generated graphics and code. Not a future to look forward to.

2. Do you also find adventure games that use 'procedural generation' to be really repetitive and gimmicky?

I don't think procedural generation was a precursor to AI. I think precursors to AI look a lot more like Cleverbot and Deep Dream, non-gaming browser tools/toys. Procedural generation can be good or bad, it really depends on how its used. I don't think it would work well in adventure games, where having a more handcrafted experience works to the other strengths of the genre.

The Desktop Adventures games are just kinda mid, and age has been even less kind to them. Desktop Adventures sits as a precursor to mobile gaming and roguelites, two genres which have been going crazy with popularity and innovation in the past decade alone.

1. Why do people use AI to make adventure games rather than using public domain and creative commons assets?

Not every kind of asset you need for your game is available as public domain or creative commons asset, when you have a specific style in mind. So if they decide to use AI generated assets instead, it is probably because they think it better matches their vision. And it might be more (time) efficient to work based on (or with) a generated asset than for example taking a public domain asset and paint over that to make it match the planned style.
And for some it also might not be a concern (ethically) at all where the asset has come from and how it has been created. They may not value an asset more when it has been created by a human, it may not be a priority for them to create a truly hand-crafted game.
Before AI, people just stole or copied assets from the internet when they needed something, cropped and used these directly. Using AI might be a lower entry level for stealing, giving them the feeling that "they have created the image by themselves" (while the AI automatically did the stealing for them invisibly in the background).

I agree with this, and would like to add more supporting points:
While there are artists who deliberately work in the open art space and they are heroes, most mass-culture public domain works are decades behind the modern day. We're getting Betty Boop in the public domain in 2026, to illustrate how behind culture the public domain is. Whether you're inspired by gorgeous pixel art, kawaii anime and otaku subculture, or the rhythm of rap music, you're straight outta luck for any of that mass culture being in the public domain. I hope you like fuckin... Kenney's game assets or some thing.

Not a dig. I like Kenney's game assets. I like Open Game Art. This is no shade to the public domain. In fact, I love it. I am a huge supporter of the public domain, in a way that puts me at odds with the socially acceptable and common opinions a lot of people hold. There's a lot I could argue if we wanted to flame war it out, but I don't lol.

I hate the copyright discussion around AI, its taken several directions I'm fundamentally at odds with. I strongly believe copyright is bad for art and bad for artists, DRM and copy protection is for losers, and art should be free. I also don't believe AI to be theft, as copying is not theft. You can figure out where the discussion frustrates me. My criticisms hits all sides, many stock asset sites are huge libraries of copyfraud. AI art itself is meant to be uncopyrightable, but people commit copyfraud with AI art every day and it pisses me off.

I wish devs could yoink assets off the Spriters Resource and sell the results for $$$, some graphics are too gorgeous to be contained to one game. The beautiful patchwork visuals of fan games is testament to that, the best fan games achieve a beautiful curation of ripped assets and original vision, and are as deserving of monetization as any project. But as my prior opinions may imply, I'm also glad they're free.

In the interests of not coming off too sympathetic to AI, I think a lot of it is poorly utilized and fugly. I do not think there's been a good looking game with large amounts of AI assets yet.

Extremely detailed answers from both Hexenwerk and Gary. Thank you. 

I'm following this discussion with great interest. I haven't voted yet, I'm still thinking about it.

Personally, I try to avoid generative AI - not for creative rights, quality or protecting the nature of art itself, though. At the same time, as far as "ethical" consumer choices go, I think this is solidly in the inconsequential category. Even if it wasn't, it's not a viable long-term strategy against the economical and ecological consequences of the way it's currently being rolled out.

I think current developments in AI tech are impressive and are occassionally used in efficient ways. In the long run, the big fish are going to get a good return on investment. Some detractors like to paint it as a frontier that's being willed into being by people in dire need of a new growth market, or some accelerationist ploy by people not all that sympathetic to humanity. I think both are mostly ideological fantasy, though some of the people pushing the tech share these fantasies with their fiercest opponents: one dreads and the other celebrates the end of society as we know it.

I dislike the combination of new growth market, disruptive tech and corresponding business practices. I think it's an excuse for the big fish to be reckless, getting the smaller fish to go along - they're going to be a nice meal before long. So in that way I do fear a bit of a bust. It don't think it's going to the same as the dot-com bubble: the money is coming from some deep pockets, including military contracts. It is hyped up beyond what it can deliver though, especially in terms of cost effienciency. I don't see it bridging that gap soon.

I think governments and other institutions with some sort of mandate of law are catching up to ways it's being applied. Ethical/legal concerns are going to result into new laws. I'm curious how this will impact things like media accessibility and state sovereignity.

I'm guessing most of the big tech dudes are going to oppose most regulations most of the time. Selectively. Once in a while it'll help if certain limitations mostly come at expense of one of your rivals. From the other side, governments are going to rely on these services - I really wonder how different forms of power will relate to one another in the near future.

I want to learn more about the artistic side of things, this discussion has had some good input in that regard.

I also wonder about the skill trade-off: how will AI literacy affect traditional forms of literacy? Not just in reading, essentially anything that requires a long process of concentration dedicated to familiarization. I suppose art falls in that category as well. Of course, they can go hand in hand and even lead to further (hyper-)specialization. Like calculators don't ruin your mental math - as long as it's already there. You'll keep doing it when it's quicker than grabbing a calculator. But for many people, that's not the case anymore.

I think what annoys me about the whole AI stuff is that there are always people who manage to make money out of something someone else has originally created. A "someone else" who is most of the time an artist who struggles to survive and make a living with their work at all.

And even with the existence of AI, there are still also those who steal (I still call it steal when they make money from it afterwards) existing illustrations (art assets). Should art be "free" for being exploited by such people? Just this week I had to contact TEMU about a copyright infringement of a work of mine. I created this for my own "IP" (worldbuilding project) which has no commercial / monetization interest at all.
And someone took it and just sold it as a print in their store without me knowing about it. I stumbled upon that accidentally.

TEMU acted quickly and removed it. I still don't understand why people are doing this. I hate it when others try to profit from others hard work (it took me months to create this) this way.

For me this has nothing to do with "make art available to everyone".

If I had generated that illustration with AI, it wouldn't matter to me. I wouldn't have any personal connection to that output.

I think what annoys me about the whole AI stuff is that there are always people who manage to make money out of something someone else has originally created. A "someone else" who is most of the time an artist who struggles to survive and make a living with their work at all.
And even with the existence of AI, there are still also those who steal (I still call it steal when they make money from it afterwards) existing illustrations (art assets). Should art be "free" for being exploited by such people? Just this week I had to contact TEMU about a copyright infringement of a work of mine. I created this for my own "IP" (worldbuilding project) which has no commercial / monetization interest at all.
And someone took it and just sold it as a print in their store without me knowing about it. I stumbled upon that accidentally.
TEMU acted quickly and removed it. I still don't understand why people are doing this. I hate it when others try to profit from others hard work (it took me months to create this) this way.
For me this has nothing to do with "make art available to everyone".
If I had generated that illustration with AI, it wouldn't matter to me. I wouldn't have any personal connection to that output.

Copyfraud sucks, and I'm very happy that TEMU took it down. That's bullshit and sucks you had to deal with that.

If it makes you feel any better, its very unlikely it made any sales. Businesses like to spam a lot of print-on-demand goods on their store, as a search engine tactic to capture as many views as possible. Upload as much as possible, in the hopes of pulling in the few niche customers. There is no cost if something fails to sell, as everything is print on demand. If you manage to get hundreds or thousands of images up, you're probably going to sell something. This is why they do it.

This is a tactic that only works in big operations doing it at scale, and done by people who don't care. And - for better or worse - this same spam but done with AI images is now ubiquitous as you'd expect.

I'm sorry you were caught in storefront spam. It sucks.

Amazon selling "Adult diaper worn by old man with a crutch" phone cases, and other inappropriate stock images nobody would ever want on a phone case, is a funny example of this phenomenon from about... oh a decade ago. From 2017. Phone cases are print on demand goods, too. Its depressing how long this has been going on.


https://www.iflscience.com/amazon-ai-des...rong-42645

Support real artists.

Support real artists.

Is an art forger using only traditional techniques a real artist? How about someone who uses AI tools available in digital processing tools? Were Renaissance painters outsourcing part of the work on their own paintings to assistants and studio hands real artists?