I'm kind of with Jen on this one, I have so many games in my library to play that don't use generative AI, I will have enough to play for the rest of my life without ever having to touch one that uses GenAI.
For now I'm happy that games on Steam and itch have an AI disclosure section, even if it's just the honor system. That's the first thing I check now. If a game makes a disclosure I immediately mentally file that game in the "never have to think about because it has no value" pile. I understand that their still room for developers to lie or mislead about this but there's still so many games these days that are making those disclosures that I do have to file a lot of games in that pile and the number of upcoming games I'm anticipating is much smaller now than it has been in the past.
However, on the coding side it's a little different. One thing that LLMs are good at is writing tests and brute forcing all cases to look for bugs. Morally, I don't like that we're propping up the valuations of companies like OpenAI and Anthropic by using ChatGPT and Claude and the societal destruction those companies are causing, but from the perspective of would that use of AI make me want to disregard a game, I would probably say no.
At some point, maybe even soon, we may get to the point where we can run models locally on our own machines that perform as well at inference as the main commercial models do today. When that happens, I won't have much of a problem for using AI for bug testing, code completion, etc. because the models will be open source, run locally on your computer so not requiring big data centers for inference and won't be any more environmentally destructive than playing games at max settings on your own device. The training will still be a problem, because that will be done on stolen work scraped from the internet and probably still require massive energy and resources so I can respect those who say they will have a problem with GenAI use of any kind.
But while we're still in this "grow users at any cost" part of the enshittification lifecycle with respect to the big AI companies, I think boycotting their tools and the "art" that is generated from them is what I would recommend to help send the message that this is not the future we want.
And from the perspective of the Adventure Game Hotspot, it would be really nice if you took a staff position of "we won't cover games on our site that make AI disclosures", or at the very least include an "AI Assisted" tag in the game details section so when I open a review and see that tag I can immediately move on.
For now I'm happy that games on Steam and itch have an AI disclosure section, even if it's just the honor system. That's the first thing I check now. If a game makes a disclosure I immediately mentally file that game in the "never have to think about because it has no value" pile. I understand that their still room for developers to lie or mislead about this but there's still so many games these days that are making those disclosures that I do have to file a lot of games in that pile and the number of upcoming games I'm anticipating is much smaller now than it has been in the past.
However, on the coding side it's a little different. One thing that LLMs are good at is writing tests and brute forcing all cases to look for bugs. Morally, I don't like that we're propping up the valuations of companies like OpenAI and Anthropic by using ChatGPT and Claude and the societal destruction those companies are causing, but from the perspective of would that use of AI make me want to disregard a game, I would probably say no.
At some point, maybe even soon, we may get to the point where we can run models locally on our own machines that perform as well at inference as the main commercial models do today. When that happens, I won't have much of a problem for using AI for bug testing, code completion, etc. because the models will be open source, run locally on your computer so not requiring big data centers for inference and won't be any more environmentally destructive than playing games at max settings on your own device. The training will still be a problem, because that will be done on stolen work scraped from the internet and probably still require massive energy and resources so I can respect those who say they will have a problem with GenAI use of any kind.
But while we're still in this "grow users at any cost" part of the enshittification lifecycle with respect to the big AI companies, I think boycotting their tools and the "art" that is generated from them is what I would recommend to help send the message that this is not the future we want.
And from the perspective of the Adventure Game Hotspot, it would be really nice if you took a staff position of "we won't cover games on our site that make AI disclosures", or at the very least include an "AI Assisted" tag in the game details section so when I open a review and see that tag I can immediately move on.
This post was last modified: Yesterday, 05:21 AM by PopPunkParty.
I think, with any luck, it’s a trend that’ll destroy itself. All indie games with AI art look equally ugly, and since they’re usually made by people lazy and inexperienced enough to rely on AI, everything beyond the graphics ends up sucking too, from the animations to the story and puzzles. I think it’ll also make people appreciate handmade, imperfect indie games more, along with games that have a unique art style or gameplay concept. The only thing missing right now is a proper filter on stores like Steam so you can hide anything made with AI.
Then there’s the fact that tokens are getting more and more expensive. I work at a mobile games company, and a year ago they told us we could use as much AI as we wanted for everything except final art. A few months later, they got an astronomical bill (the company has offices all over the world and more than 2,000 employees), so they capped token spending for each seat at €1,000/$1,000 a month. Last week, just three or four months later, they told us they’d cut the cap to €500 a month, and some people on the tech team burn through that in a single week.
One approach people could take is to just scroll past when someone shows off their AI-made “dream project.” Don’t comment, don’t download the demo, and definitely don’t pay for it. I’m not calling for a backlash, since the people behind these projects usually mean well. I’m just saying to don’t have to give their projects any attention or visibility.
Then there’s the fact that tokens are getting more and more expensive. I work at a mobile games company, and a year ago they told us we could use as much AI as we wanted for everything except final art. A few months later, they got an astronomical bill (the company has offices all over the world and more than 2,000 employees), so they capped token spending for each seat at €1,000/$1,000 a month. Last week, just three or four months later, they told us they’d cut the cap to €500 a month, and some people on the tech team burn through that in a single week.
One approach people could take is to just scroll past when someone shows off their AI-made “dream project.” Don’t comment, don’t download the demo, and definitely don’t pay for it. I’m not calling for a backlash, since the people behind these projects usually mean well. I’m just saying to don’t have to give their projects any attention or visibility.
This post was last modified: Yesterday, 08:11 AM by Karellen Games.
Personally, I'm not bothered about how AI is / was trained. People use the internet to freely expose their work all the time and no one who does that can expect to be safe from having it copied by anyone, including LLMs. The more successful you are the more you stand out and the more you'll be copied. Economies develop over time and things that were once controlled by a few big companies get manipulated by other big companies (like Apple Music, for example) and I don't see AI companies being different to that.
I'm also in two minds about use of AI in a creative process. The only thing I really use AI for is helping with coding because I'm not very good at it. But, if I wasn't good at drawing I'd probably want help with that instead. In either case, someone is potentially losing out on being paid to help me. That said - it's currently usually fairly easy to spot AI art and, yes, it does tend to look generic and slightly disturbing.
The last point is the most important for me - the environmental cost of running this world wide facility is bound to be massive, and it's coming at a point when things are already bad. That would be the real reason to turn it off, as far as I'm concerned.
I'm also in two minds about use of AI in a creative process. The only thing I really use AI for is helping with coding because I'm not very good at it. But, if I wasn't good at drawing I'd probably want help with that instead. In either case, someone is potentially losing out on being paid to help me. That said - it's currently usually fairly easy to spot AI art and, yes, it does tend to look generic and slightly disturbing.
The last point is the most important for me - the environmental cost of running this world wide facility is bound to be massive, and it's coming at a point when things are already bad. That would be the real reason to turn it off, as far as I'm concerned.