So many companies and individuals using AI and are like, here's our (sloppy) product! Whether it's a game or program or just a blog post. But it's not theirs. They didn't create it. They can't claim copyright over it. Even if they change it a little bit, I don't get how that somehow is transformative enough to grant any protection. So unless how copyright works drastically changes, I don't understand what they think is going to happen, at least from a perspective of a person or company who is doing this to make money.
I think you're right in the scenario there was no human involvement in design. my guess which may be naive is that the tool use will be mixed. You may see gen AI used more for placeholder assets for example. I also think the landscape is shifting to the point codegen is acceptable in those circles.
If there is mixed usage of AI and human design then like the article said those components generated likely aren't copyrightable but the specific way you cobbled them together could be and that wouldn't need much change it seems. I'd like to follow the revenue anyway, last article I read was from about 6 months ago that steam AI-gen declarations accounted for almost 700million in revenue but it was unclear then to what extent ai tools were used (https://www.gamesradar.com/games/steam-games-that-openly-use...)
> You may see gen AI used more for placeholder assets for example.
Is it worth the risk, though? AI-generated content has become good enough that it isn't always immediately obvious that it is AI-generated. This means some poorly-paid QA tester isn't guaranteed to flag it, so it might very well end up slipping into the final release - as has already happened multiple times now.
Need a placeholder four-headed dog? Grab a dog model from any random library, cut off its head, add four Blender Cubes on stalks, and you're done. Less than five minutes of work and it will stand out as "obviously a placeholder" during QA.
Hmm regarding quality i get mixed signals from this and other comments. To be honest i don't say you're wrong about that risk being real but it's clear reading the other person's comments on quality being actually poor with assets that there is some assessment in both quality and usage that I don't fully understand. The placeholder use is indeed based on my reading of studios own accounts, but if you say it is problematic, I can't argue that as I have no other experience here and tbh if a game is fun I'd take the dog with no head and cubes too.
I would say the quality is usually too poor to ship, but good enough that it can slip through unnoticed.
It's like a web developer using their own photo as a placeholder for the CEO on an "About Us" page: seems like a fun little in-joke, until it accidentally ships and you get an angry phone call from your boss's boss. It's better to do the boring thing and play it save by using a big red cross with "REPLACE ME" on it so you'll definitely catch it during the final review.
Its happening in other industries as well, as legal and contract management teams have stepped up their maturity in considering these things.
I work in Consulting and we're seeing this across our industry as well. Multiple clauses popping up around no or limited AI use, which comes from those organisations being risk averse, through not knowing how to deal with the reputational risk if AI were to cause an issue.
We see some organisations being very strict on no AI usage, to other organisations naming particular platforms or particular classifications of information, which is a smart way to go about it.
I suspect that won't last long for a number of reasons
- AI is already capable of creating some assets that are indistinguishable from human made ones. Anti-AI sentiment is pointless if they can't tell the difference.
- devs in Countries that care a lot less about the legal stuff are going to be cranking out and studios will want in on it.
- If the primary blocker is legal someone will find a way to get around it and provide a model/service/workflow that allows a game with AI assets to claim copyright on said assets.
- Games are literally one of the absolute best uses for this tech. Someone is going to figure out a way to leverage this stuff as a first class function of the gameplay. Player directed assets or procedural world creation, dynamic quests like never before seen. They're going to hit the nail on the head and produce a banger with this tech and the landscape is never going to be the same again.
This typically doesn't matter at all when it comes to sales. Take the fervent Hogwarts Legacy boycott that was all over the internet for a very long time. That ended with checks notes the first game to beat sales of a Call of Duty or a Rockstar game in 14 years and 40 million+ copies sold. My guess is a similar thing will happen with EA being Saudi owned as well.
It's the vocal minority that will care if AI is in a game as well. So long as the game is fun very few people will even have the ability to know let alone care.
> Games are literally one of the absolute best uses for this tech. Someone is going to figure out a way to leverage this stuff as a first class function of the gameplay. Player directed assets or procedural world creation, dynamic quests like never before seen. They're going to hit the nail on the head and produce a banger with this tech and the landscape is never going to be the same again.
I see the AI boosters say this over and over again. People do not want it. Games have tried the whole dynamic NPC stuff along with generated quests and it's immersion breaking and more importantly, not fun.
People definitely want it. I can say I want to see it happen. Also people don't know to want what they don't know exists.
There just hasn't been the Edison light-bulb/Watt's steam engine/Apple iPhone moment yet. I'm not even sure generative AI is capable of it yet but I see the potential there. While not all dynamic games are "not fun" (lotta people really like Rimworld) that kind of content tends to feel flat, like the never-ending quests in Skyrim. But I think it's easy to squint and see how those shallow Skyrim quests could become far more interesting with a system that was capable of basically writing a story as you play.
IMHO It's only a matter of time before someone gets it right.
i think good dynamic quests will eventually be a thing but not coming from gen AI or old school complicated rule books. if you generate the whole thing with a LLM you get a lot of results that make no sense. if you go with a deterministic rule based generator or a constraint solver you get low diversity because the rules are fixed.
my solution is a hybrid approach using a machine learning model trained to estimate player ratings for a quest script. you have a simple RNG that generates a ton of different script, the model picks out the ones that are likely to be fun and you give one of them to the player.
the meta might be using all the methods (maybe not LLMs) together to get a result thats fun, logically consistent and even fits into a larger narrative. like tracking NPC relationships to assign a target for combat missions, coming up with extra plot twists along the way then using ML to rate all the generated options.
I've worked in the industry and agree with the parent. This person hasn't worked in the industry.
AI is a bad for the industry for a few reasons:
Players hate it. You think the vitriol over "asset store flips" is bad? AI is seen by players as lazy and low quality.
Copyright is uncertain. IP matters a lot to studios. Until that's settled, it's caution.
Art is never the bottleneck in games. It's why so much DLC is cosmetic items. Artists finish and gotta be making more stuff, but the rest of the studio is still powering through the final feature list.
Games are perceived as expressive, artistic, and creative works. Everyone is working together to build something with their human identity stamped in it. This is what we're making, together. The moment you invite slop in, the product isn't art anymore. It's just a product. Sure, maybe shovelware titles will use AI, but audiences will know.
The quality of AI assets is pretty mid. They feel samey and tacky. No one wants to play games that feel like the average game. They want games that are novel or familiar in unusual ways.
It's expensive to even try. Studios run on razor thin margins and tight deadlines. You want to spend two months on some spike to see if AI might pay off? Good luck, we'll go bankrupt.
Production schedules have orders, estimates, and processes. People do stuff in a particular order and timeframe. Disrupting a machine that works and replacing it with a machine that may work is a risk. Best case it pays off, worse case it's development hell, worst case it's a canceled game.
Worker solidarity. It's thin in game dev, but it's growing. There's a sense that the top is keeping the profit, and AI is a management driven tool to keep the worker suppressed. It's not universal, but it's present.
There's probably a dozen more reasons. Cost of tokens, tracking provenance (did this artist copy another artists work is really really important to be able to prove), high degree of change in AI ecosystem right now disrupting tool chains, it's uncertain how the market will react in the long term to games with AI assets (gamedev is a conservative industry). Etc.
Thanks. It's hard as an outsider to evaluate the counters when you have nothing but vague criticism of the comment. I think you laid out the reasons clearly, thanks for taking the effort.
Making games is hard! All the flaws and limitations of AI can't be handwaved away like they can with business software, the CRUD app shovelware that gets posted here daily. Or with passages of text that look valid on the surface. There's a reason Amazon and Google failed to get meaningfully into gaming, it's because it's something that can't be bought. The more deeply hands on you've been with making games, the less optimistic you're likely to be about automating game dev. Surely individual slop assets are making their way into a game at a low level but that's different than the vague grand vision of "Player directed assets or procedural world creation, dynamic quests" etc.
> AI is already capable of creating some assets that are indistinguishable from human made ones.
Some assets, yes. But gamers are already losing their mind over single-digit number of AI assets (allegedly) accidentally left into a game. Even when 99% of assets are indistinguishable, you'll still get crucified by the anti-AI gamer mob for the 1% where they can.
> Anti-AI sentiment is pointless if they can't tell the difference.
The gamer crowd already hates it when you are caught reusing human-made stock assets. They hate any perceived laziness. AI is just another addition to this.
> devs in Countries that care a lot less about the legal stuff are going to be cranking out and studios will want in on it.
The studios have already been trying, hence the aforementioned backlash. The studios might want in on it due to higher productivity, but the publishers are against it because the negative PR isn't worth it.
> Someone is going to figure out a way to leverage this stuff as a first class function of the gameplay. Player directed assets or procedural world creation, dynamic quests like never before seen
Been there, done that, got the t-shirt. Choose-your-own-adventure did not kill the traditional book, Open-world and sandbox games did not kill linear storytelling. Procedural worlds universally feel empty and repetitive because they are.
People recognize a "kill 15 boars" quest when they see one: killing 10 wolves or collecting 20 unicorn horns doesn't change the fundamental structure of what it is. No matter the window dressing, people will get tired of repetitive quests generated from the same prompt. What makes a game like, say, World of Warcraft worth playing is that the thousands of quests tie together into a coherent overarching plot, where you and your friends work together for hundreds of hours on everything from disrupting supply lines, to smoking out infiltrators, to helping a kid who lost his parents during an attack, to eventually defeating the Big Bad Evil. You simply cannot do this with a one-off personalized storyline which makes up additional things as you play it.
Besides, apps like Infinite Worlds[0] are already a thing: it lets people create their own story world in which other people can experience AI-driven text adventures. Sounds great, right? Turns out a handful of people do indeed like it, but for the most part the communities for both readers and writers are banishing it to special AI sections because most people dislike it and are sick of it flooding the rest of the platform.
There are already enough books and games out there to meet everyone's desires. AI-driven storytelling does not meaningfully tap any market of unexplored narratives, while at the same time sacrificing story quality and any sense of shared experience. AI storytelling won't replace traditional storytelling in the same way that AI waifus aren't going to replace real-world (platonic or romantic) relationships.
So many companies and individuals using AI and are like, here's our (sloppy) product! Whether it's a game or program or just a blog post. But it's not theirs. They didn't create it. They can't claim copyright over it. Even if they change it a little bit, I don't get how that somehow is transformative enough to grant any protection. So unless how copyright works drastically changes, I don't understand what they think is going to happen, at least from a perspective of a person or company who is doing this to make money.
I think you're right in the scenario there was no human involvement in design. my guess which may be naive is that the tool use will be mixed. You may see gen AI used more for placeholder assets for example. I also think the landscape is shifting to the point codegen is acceptable in those circles.
If there is mixed usage of AI and human design then like the article said those components generated likely aren't copyrightable but the specific way you cobbled them together could be and that wouldn't need much change it seems. I'd like to follow the revenue anyway, last article I read was from about 6 months ago that steam AI-gen declarations accounted for almost 700million in revenue but it was unclear then to what extent ai tools were used (https://www.gamesradar.com/games/steam-games-that-openly-use...)
> You may see gen AI used more for placeholder assets for example.
Is it worth the risk, though? AI-generated content has become good enough that it isn't always immediately obvious that it is AI-generated. This means some poorly-paid QA tester isn't guaranteed to flag it, so it might very well end up slipping into the final release - as has already happened multiple times now.
Need a placeholder four-headed dog? Grab a dog model from any random library, cut off its head, add four Blender Cubes on stalks, and you're done. Less than five minutes of work and it will stand out as "obviously a placeholder" during QA.
Hmm regarding quality i get mixed signals from this and other comments. To be honest i don't say you're wrong about that risk being real but it's clear reading the other person's comments on quality being actually poor with assets that there is some assessment in both quality and usage that I don't fully understand. The placeholder use is indeed based on my reading of studios own accounts, but if you say it is problematic, I can't argue that as I have no other experience here and tbh if a game is fun I'd take the dog with no head and cubes too.
I would say the quality is usually too poor to ship, but good enough that it can slip through unnoticed.
It's like a web developer using their own photo as a placeholder for the CEO on an "About Us" page: seems like a fun little in-joke, until it accidentally ships and you get an angry phone call from your boss's boss. It's better to do the boring thing and play it save by using a big red cross with "REPLACE ME" on it so you'll definitely catch it during the final review.
Its happening in other industries as well, as legal and contract management teams have stepped up their maturity in considering these things.
I work in Consulting and we're seeing this across our industry as well. Multiple clauses popping up around no or limited AI use, which comes from those organisations being risk averse, through not knowing how to deal with the reputational risk if AI were to cause an issue.
We see some organisations being very strict on no AI usage, to other organisations naming particular platforms or particular classifications of information, which is a smart way to go about it.
I suspect that won't last long for a number of reasons
- AI is already capable of creating some assets that are indistinguishable from human made ones. Anti-AI sentiment is pointless if they can't tell the difference.
- devs in Countries that care a lot less about the legal stuff are going to be cranking out and studios will want in on it.
- If the primary blocker is legal someone will find a way to get around it and provide a model/service/workflow that allows a game with AI assets to claim copyright on said assets.
- Games are literally one of the absolute best uses for this tech. Someone is going to figure out a way to leverage this stuff as a first class function of the gameplay. Player directed assets or procedural world creation, dynamic quests like never before seen. They're going to hit the nail on the head and produce a banger with this tech and the landscape is never going to be the same again.
> Anti-AI sentiment is pointless if they can't tell the difference.
I disagree, you can object on moral grounds of which there's a few angles that could take.
This typically doesn't matter at all when it comes to sales. Take the fervent Hogwarts Legacy boycott that was all over the internet for a very long time. That ended with checks notes the first game to beat sales of a Call of Duty or a Rockstar game in 14 years and 40 million+ copies sold. My guess is a similar thing will happen with EA being Saudi owned as well.
Perhaps you are making the mistake of confusing a widespread boycott with a vocal minority.
It's the vocal minority that will care if AI is in a game as well. So long as the game is fun very few people will even have the ability to know let alone care.
Much of the games industry is driven by 12 year olds who are not discerners of taste but driven by admen.
But rather than regulate the social media industry driving the 12-year old industry, they'd rather push all the under 16 social media bans.
Regardless, the people who drive videos games are not adults and are unlikely to factor into the antiai
> Games are literally one of the absolute best uses for this tech. Someone is going to figure out a way to leverage this stuff as a first class function of the gameplay. Player directed assets or procedural world creation, dynamic quests like never before seen. They're going to hit the nail on the head and produce a banger with this tech and the landscape is never going to be the same again.
I see the AI boosters say this over and over again. People do not want it. Games have tried the whole dynamic NPC stuff along with generated quests and it's immersion breaking and more importantly, not fun.
People definitely want it. I can say I want to see it happen. Also people don't know to want what they don't know exists.
There just hasn't been the Edison light-bulb/Watt's steam engine/Apple iPhone moment yet. I'm not even sure generative AI is capable of it yet but I see the potential there. While not all dynamic games are "not fun" (lotta people really like Rimworld) that kind of content tends to feel flat, like the never-ending quests in Skyrim. But I think it's easy to squint and see how those shallow Skyrim quests could become far more interesting with a system that was capable of basically writing a story as you play.
IMHO It's only a matter of time before someone gets it right.
i think good dynamic quests will eventually be a thing but not coming from gen AI or old school complicated rule books. if you generate the whole thing with a LLM you get a lot of results that make no sense. if you go with a deterministic rule based generator or a constraint solver you get low diversity because the rules are fixed.
my solution is a hybrid approach using a machine learning model trained to estimate player ratings for a quest script. you have a simple RNG that generates a ton of different script, the model picks out the ones that are likely to be fun and you give one of them to the player.
the meta might be using all the methods (maybe not LLMs) together to get a result thats fun, logically consistent and even fits into a larger narrative. like tracking NPC relationships to assign a target for combat missions, coming up with extra plot twists along the way then using ML to rate all the generated options.
I'm going to guess you've never worked in game dev and I'd be shocked if I was wrong.
Can you add a bit more? It would be great to hear the actual counters of how that industry thinks.
I've worked in the industry and agree with the parent. This person hasn't worked in the industry.
AI is a bad for the industry for a few reasons:
Players hate it. You think the vitriol over "asset store flips" is bad? AI is seen by players as lazy and low quality.
Copyright is uncertain. IP matters a lot to studios. Until that's settled, it's caution.
Art is never the bottleneck in games. It's why so much DLC is cosmetic items. Artists finish and gotta be making more stuff, but the rest of the studio is still powering through the final feature list.
Games are perceived as expressive, artistic, and creative works. Everyone is working together to build something with their human identity stamped in it. This is what we're making, together. The moment you invite slop in, the product isn't art anymore. It's just a product. Sure, maybe shovelware titles will use AI, but audiences will know.
The quality of AI assets is pretty mid. They feel samey and tacky. No one wants to play games that feel like the average game. They want games that are novel or familiar in unusual ways.
It's expensive to even try. Studios run on razor thin margins and tight deadlines. You want to spend two months on some spike to see if AI might pay off? Good luck, we'll go bankrupt.
Production schedules have orders, estimates, and processes. People do stuff in a particular order and timeframe. Disrupting a machine that works and replacing it with a machine that may work is a risk. Best case it pays off, worse case it's development hell, worst case it's a canceled game.
Worker solidarity. It's thin in game dev, but it's growing. There's a sense that the top is keeping the profit, and AI is a management driven tool to keep the worker suppressed. It's not universal, but it's present.
There's probably a dozen more reasons. Cost of tokens, tracking provenance (did this artist copy another artists work is really really important to be able to prove), high degree of change in AI ecosystem right now disrupting tool chains, it's uncertain how the market will react in the long term to games with AI assets (gamedev is a conservative industry). Etc.
Thanks. It's hard as an outsider to evaluate the counters when you have nothing but vague criticism of the comment. I think you laid out the reasons clearly, thanks for taking the effort.
Making games is hard! All the flaws and limitations of AI can't be handwaved away like they can with business software, the CRUD app shovelware that gets posted here daily. Or with passages of text that look valid on the surface. There's a reason Amazon and Google failed to get meaningfully into gaming, it's because it's something that can't be bought. The more deeply hands on you've been with making games, the less optimistic you're likely to be about automating game dev. Surely individual slop assets are making their way into a game at a low level but that's different than the vague grand vision of "Player directed assets or procedural world creation, dynamic quests" etc.
> AI is already capable of creating some assets that are indistinguishable from human made ones.
Some assets, yes. But gamers are already losing their mind over single-digit number of AI assets (allegedly) accidentally left into a game. Even when 99% of assets are indistinguishable, you'll still get crucified by the anti-AI gamer mob for the 1% where they can.
> Anti-AI sentiment is pointless if they can't tell the difference.
The gamer crowd already hates it when you are caught reusing human-made stock assets. They hate any perceived laziness. AI is just another addition to this.
> devs in Countries that care a lot less about the legal stuff are going to be cranking out and studios will want in on it.
The studios have already been trying, hence the aforementioned backlash. The studios might want in on it due to higher productivity, but the publishers are against it because the negative PR isn't worth it.
> Someone is going to figure out a way to leverage this stuff as a first class function of the gameplay. Player directed assets or procedural world creation, dynamic quests like never before seen
Been there, done that, got the t-shirt. Choose-your-own-adventure did not kill the traditional book, Open-world and sandbox games did not kill linear storytelling. Procedural worlds universally feel empty and repetitive because they are.
People recognize a "kill 15 boars" quest when they see one: killing 10 wolves or collecting 20 unicorn horns doesn't change the fundamental structure of what it is. No matter the window dressing, people will get tired of repetitive quests generated from the same prompt. What makes a game like, say, World of Warcraft worth playing is that the thousands of quests tie together into a coherent overarching plot, where you and your friends work together for hundreds of hours on everything from disrupting supply lines, to smoking out infiltrators, to helping a kid who lost his parents during an attack, to eventually defeating the Big Bad Evil. You simply cannot do this with a one-off personalized storyline which makes up additional things as you play it.
Besides, apps like Infinite Worlds[0] are already a thing: it lets people create their own story world in which other people can experience AI-driven text adventures. Sounds great, right? Turns out a handful of people do indeed like it, but for the most part the communities for both readers and writers are banishing it to special AI sections because most people dislike it and are sick of it flooding the rest of the platform.
There are already enough books and games out there to meet everyone's desires. AI-driven storytelling does not meaningfully tap any market of unexplored narratives, while at the same time sacrificing story quality and any sense of shared experience. AI storytelling won't replace traditional storytelling in the same way that AI waifus aren't going to replace real-world (platonic or romantic) relationships.
[0]: https://infiniteworlds.app/