You run the chances of being perhaps more easily targeted by self-righteous assholes, sacrificing for the common good you hope to also benefit from. If you appreciate that being able to separately browse between AI and non-AI works is useful, especially without particular moral judgement, and recognize this capability as socially facilitated, then the way to cultivate that behavior is to take the leap if you can yourself.
As long as people are on the lookout for AI stuff with their eyes in the red, this won't work. If you invert the mechanism and tell people hey my shit is AI, you take the wind out their sails, and directly disprove that you're some monster trying to trick or hurt them. It's easily avoided, so that narrative doesn't work anymore. Less ground for the villification to work on their part, as you can no longer be the cartoon villian they want to think you are.
Not the first platform with this feature by far either.
It's basically an honor system. The point is exactly to show that you're willing to be (somewhat) vulnerable. You can only demonstrate that you're not misusing people's trust, if there is trust to misuse. As long as AI usage is treated as betrayal, this is what remains. https://en.wikipedia.org/wiki/Honor_system
It's very distressing that you can't even fathom people would self-disclose. You clearly understand that a lot of people would refrain from purchasing if they knew...
If you can develop something that doesn't appear to be obviously developed with LLMs, what is the point of disclosing it? It serves no positive purpose for the developer.
Ethics? Respect for your consumer? Can you imagine if the food industry operated under your reasoning? If the consumer can't tell that the "ground beef" is rat meat|dog meat|human flesh, what purpose does it serve to disclose it? It serves no purpose for the manufacturer.
I'm sure a lot of people wouldn't care, but you "know" some consumers don't want it. At least give them the choice.
There's a massive difference between something you ingest into your body, and a silly indie video game you play on Itch. One can affect your health, the other effects only emotions
How about feeding a vegan (for ethical reasons) real meat without their knowledge? Is that a more fitting example? Or products built with child/slave labor?[1]
Look at the backlash received when it came out that Clair Obscur: Expedition 33 used AI assets[2]. By all rights, it was a good game, even won GOTY (though later stripped for the AI usage). There are a lot of people that would not have bought it if they knew it used GenAI, and there's a lot of people that still would have.
Bottom line: Consumers deserve the right to choose.
Yes, do look at the Clair Obscur frenzy. It was over placeholder textures slipping through. A newspaper clipping on a post.
This isn't about feeding the ethical vegan meat. It's about the farm workers growing the vegetables needing to be vegan themselves. If customers wish to limit their buying to only vegan food grown by vegans, that's certainly their right to choose, and those companies are free to label their products as such.
"They said" it was only placeholder textures. Users reported dialog seemed AI generated. Only the developers know the truth. Either way, enough people were put off by it's inclusion (in any form) that it lost GOTY and probably a lot of sales.
Regarding your "solution", that may well be a good solution. Itch.io could use a similar search filter to only return results that have a `no-gen-ai` label or similar. But unless that sort of label became an actually regulated term, I suspect if (when) the filter effected sales people would start adding it whether it were true or not. Look at how food packaging has all sorts of creative interpretations of "organic" or similar terms.
On that same thought, does the "this contains GMO ingredients" label of foods unfairly impact manufacturers? How hard is it to just responsibly disclose how something is made?
This whole argument is "you know people don't want it, and you don't want to tell people it's there because then they won't buy your thing".
> "They said" it was only placeholder textures. Users reported dialog seemed AI generated. Only the developers know the truth. Either way, enough people were put off by it's inclusion (in any form) that it lost GOTY and probably a lot of sales.
Hence a purity test that can't be passed. Did the dialog seem AI generated, or just stiff? What kind of evidence would be enough? No proof would ever be enough if you don't trust the develop when they say it was placeholder.
> Regarding your "solution"
I am not sure why you feel the need to put scare quotes around this version, but having an "this was made with AI" tag isn't also a "solution". It's literally the same purity test.
> this contains GMO ingredients
I think the comparison to GMOs is an interesting one because that one's a big nothing burger. People just have an irrational fear of GMO because it's not "natural".
It's the same idiocy that leads people to religiously check for nitrites on their cured meat packaging while gladly eating it if it contains celery juice (hint they're using the celery juice for it's natural occurring nitrite).
This is a hard problem to solve that always existed but is exacerbated by the barrier to entry dropping to the ground.
Really itch is kinda hitting what the mobile markets have been dealing with since long before LLMs, those markets were awash with clones and low quality even outright scammy crap.
Though maybe a bit harder for Itch since their whole shtick is being a low barrier place for people to put their game. Arguably the platform has always had slop, but I would categorize the majority as "honest slop", a person who maybe isn't any good at making games, or doesn't follow through actually finishing their game putting something up there.
anyone old enough to remember Newgrounds and the era of flash knows that loads of flash games were awful. but you would get a more normal Pareto distribution. Ai has lowered the bar so low that ratio becomes skewed. like 1:99 instead of 80:20 with the flood.
Banning AI seems like a non starter, you can make "honest slop" with ai, you can make even good games with it. Self tagging is basically saying "i don't want anyone to play my game". an honest game with AI will get thrown into the pit with the total crap. it's death.
chewing up real people in filters or auto-mod just makes the problem worse. Real people give up on the platform because they were burned, only the AI slop persists.
I don't know how you retain an open and friendly platform that's accepting of honest slop but can stem the tide of genuine slop.
The issue with AI slop is not so much the slop itself (there was always slop) but the amount of slop. When people are producing slop they are limited by time and ability, so the slop can go from mildly annoying to quite bad, but no more. However when it's entirely machine produced, you're literally drowned in slop if nothing is done. It's like the difference between being crapped on by a pigeon and falling in a manure pool
I mean its refreshing to read the opinions of people on the opposite side of this.. since my own feelings about gen ai are mixed. preferable to the hype
However some of the comments.. are a bit absolutist. And I don't think they are sustainable long term. I absolutely agree that the deluge of slop is a problem.
Why anyone would willingly do this is beyond me. It's like putting a sticker on your own back that says "Kick me"
Because hopefully soon the alternative will be getting banned from itch.
How is anyone going to prove it?
You can tell Claude few-shots by eye; and asking for verification, such as a few screenshots isn’t unreasonable.
You don’t have to strive for perfection. Just get rid of the worst examples, and you’ll have a deterrent effect.
You remove the obvious stuff and let the rest slide (for now).
It's called civility.
You run the chances of being perhaps more easily targeted by self-righteous assholes, sacrificing for the common good you hope to also benefit from. If you appreciate that being able to separately browse between AI and non-AI works is useful, especially without particular moral judgement, and recognize this capability as socially facilitated, then the way to cultivate that behavior is to take the leap if you can yourself.
As long as people are on the lookout for AI stuff with their eyes in the red, this won't work. If you invert the mechanism and tell people hey my shit is AI, you take the wind out their sails, and directly disprove that you're some monster trying to trick or hurt them. It's easily avoided, so that narrative doesn't work anymore. Less ground for the villification to work on their part, as you can no longer be the cartoon villian they want to think you are.
Not the first platform with this feature by far either.
It's basically an honor system. The point is exactly to show that you're willing to be (somewhat) vulnerable. You can only demonstrate that you're not misusing people's trust, if there is trust to misuse. As long as AI usage is treated as betrayal, this is what remains. https://en.wikipedia.org/wiki/Honor_system
It's very distressing that you can't even fathom people would self-disclose. You clearly understand that a lot of people would refrain from purchasing if they knew...
If you can develop something that doesn't appear to be obviously developed with LLMs, what is the point of disclosing it? It serves no positive purpose for the developer.
Ethics? Respect for your consumer? Can you imagine if the food industry operated under your reasoning? If the consumer can't tell that the "ground beef" is rat meat|dog meat|human flesh, what purpose does it serve to disclose it? It serves no purpose for the manufacturer.
I'm sure a lot of people wouldn't care, but you "know" some consumers don't want it. At least give them the choice.
There's a massive difference between something you ingest into your body, and a silly indie video game you play on Itch. One can affect your health, the other effects only emotions
How about feeding a vegan (for ethical reasons) real meat without their knowledge? Is that a more fitting example? Or products built with child/slave labor?[1]
Look at the backlash received when it came out that Clair Obscur: Expedition 33 used AI assets[2]. By all rights, it was a good game, even won GOTY (though later stripped for the AI usage). There are a lot of people that would not have bought it if they knew it used GenAI, and there's a lot of people that still would have.
Bottom line: Consumers deserve the right to choose.
[1]: https://www.bbc.com/news/business-53481253
[2]: https://www.ign.com/articles/indie-game-awards-strips-clair-...
Yes, do look at the Clair Obscur frenzy. It was over placeholder textures slipping through. A newspaper clipping on a post.
This isn't about feeding the ethical vegan meat. It's about the farm workers growing the vegetables needing to be vegan themselves. If customers wish to limit their buying to only vegan food grown by vegans, that's certainly their right to choose, and those companies are free to label their products as such.
"They said" it was only placeholder textures. Users reported dialog seemed AI generated. Only the developers know the truth. Either way, enough people were put off by it's inclusion (in any form) that it lost GOTY and probably a lot of sales.
Regarding your "solution", that may well be a good solution. Itch.io could use a similar search filter to only return results that have a `no-gen-ai` label or similar. But unless that sort of label became an actually regulated term, I suspect if (when) the filter effected sales people would start adding it whether it were true or not. Look at how food packaging has all sorts of creative interpretations of "organic" or similar terms.
On that same thought, does the "this contains GMO ingredients" label of foods unfairly impact manufacturers? How hard is it to just responsibly disclose how something is made?
This whole argument is "you know people don't want it, and you don't want to tell people it's there because then they won't buy your thing".
> "They said" it was only placeholder textures. Users reported dialog seemed AI generated. Only the developers know the truth. Either way, enough people were put off by it's inclusion (in any form) that it lost GOTY and probably a lot of sales.
Hence a purity test that can't be passed. Did the dialog seem AI generated, or just stiff? What kind of evidence would be enough? No proof would ever be enough if you don't trust the develop when they say it was placeholder.
> Regarding your "solution"
I am not sure why you feel the need to put scare quotes around this version, but having an "this was made with AI" tag isn't also a "solution". It's literally the same purity test.
> this contains GMO ingredients
I think the comparison to GMOs is an interesting one because that one's a big nothing burger. People just have an irrational fear of GMO because it's not "natural".
It's the same idiocy that leads people to religiously check for nitrites on their cured meat packaging while gladly eating it if it contains celery juice (hint they're using the celery juice for it's natural occurring nitrite).
This is a hard problem to solve that always existed but is exacerbated by the barrier to entry dropping to the ground.
Really itch is kinda hitting what the mobile markets have been dealing with since long before LLMs, those markets were awash with clones and low quality even outright scammy crap.
Though maybe a bit harder for Itch since their whole shtick is being a low barrier place for people to put their game. Arguably the platform has always had slop, but I would categorize the majority as "honest slop", a person who maybe isn't any good at making games, or doesn't follow through actually finishing their game putting something up there.
anyone old enough to remember Newgrounds and the era of flash knows that loads of flash games were awful. but you would get a more normal Pareto distribution. Ai has lowered the bar so low that ratio becomes skewed. like 1:99 instead of 80:20 with the flood.
Banning AI seems like a non starter, you can make "honest slop" with ai, you can make even good games with it. Self tagging is basically saying "i don't want anyone to play my game". an honest game with AI will get thrown into the pit with the total crap. it's death.
chewing up real people in filters or auto-mod just makes the problem worse. Real people give up on the platform because they were burned, only the AI slop persists.
I don't know how you retain an open and friendly platform that's accepting of honest slop but can stem the tide of genuine slop.
The issue with AI slop is not so much the slop itself (there was always slop) but the amount of slop. When people are producing slop they are limited by time and ability, so the slop can go from mildly annoying to quite bad, but no more. However when it's entirely machine produced, you're literally drowned in slop if nothing is done. It's like the difference between being crapped on by a pigeon and falling in a manure pool
The comments here are like a Bizarro world version of HN. Replace pro gen ai sentiment with negative
This is what they're dealing with: https://i.imgur.com/w25SmDR.jpeg
I mean its refreshing to read the opinions of people on the opposite side of this.. since my own feelings about gen ai are mixed. preferable to the hype
However some of the comments.. are a bit absolutist. And I don't think they are sustainable long term. I absolutely agree that the deluge of slop is a problem.