Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

From the submitted article:

> "When it was submitted for consideration, representatives of Sandfall Interactive agreed that no gen AI was used in the development of Clair Obscur: Expedition 33. In light of Sandfall Interactive confirming the use of gen AI art in production on the day of the Indie Game Awards 2025 premiere, this does disqualify Clair Obscur: Expedition 33 from its nomination."

Whatever placeholder you use is part of your development process, whether it ships or not. Saying they used none when they did is not cool and rightfully makes one wonder what other uses they may be hiding (or “forgetting”). Especially when apparently they only clarified it when it was too late.

I can understand the Indie Game Awards preferring to act now. Had they done nothing, they would have been criticised too by other people for not enforcing their own rules. They no doubt would’ve preferred to not have to deal with the controversy. Surely this wasn’t an easy decision for them, as it ruined their ceremony.

We’re all bystanders here with very little information, so I’d refrain from using unserious expressions like “witch hunt”, especially considering their more recent connotations (i.e. in modern times, “witch hunt” is most often used by bad actors attempting to discredit legitimate investigations).





> Whatever placeholder you use is part of your development process, whether it ships or not. Saying they used none when they did is not cool and rightfully makes one wonder what other uses they may be hiding (or “forgetting”). Especially when apparently they only clarified it when it was too late.

If it was malicious they wouldn't say a word. They probably interpreted the rule as "nothing in shipped game is AI" (which is reasonable interpreteation IMO), they implemented policy to replace any asset made by AI and just missed some texture.

Also the term was pretty vague, like, is using automatic lipsync forbidden ? That's pretty much generative AI, just the result is not picture but a sequence of movements.


> If it was malicious they wouldn't say a word.

They didn’t have a choice, it was obvious it was AI. They might still have other places where they used it but it’s harder to notice.

https://news.ycombinator.com/item?id=46344279


Malicious or not, they didn't follow the rules, and admitted as much. So why is it a problem they lose the award?

You're only saying that because it's a witch-hunt. These sorts of oversights are routinely made and forgiven daily at all levels of society. Citi regularly sends millions of dollars to wrong accounts and then says "my bad, give it back" and no one bats an eye.

But some ai gen'd placeholders beak through and suddenly we're all about punishing an oversight?

That's the definition of a witch hunt, and it's past time we admit it.


It's literally not the definition of a witch hunt. They weren't hunting for AI use in E33 or other nominees. Sandfall was stripped of the award because it admitted to using AI after the fact, despite pledging that no AI tools had been used *at any point of the development process.

Sandfall opened themselves up to this when they accepted the nomination knowing the rules for this particular award prohibited AI use of any kind.

Note that there is also now a discussion about how much AI was used to design the enemies as well, given the rather bizarre appearances of many of them enemies in the latter portions of the game.


> That's the definition of a witch hunt

You gave examples of a witch hunt. Not definition.


> Saying they used none when they did is not cool and rightfully makes one wonder what other uses they may be hiding (or “forgetting”). Especially when apparently they only clarified it when it was too late.

The article where Meurisse admitted to using AI in the pipeline is from April. You're implying a level of dishonesty that clearly isn't there.


Conceding our vocabulary to bad actors is Orwellian.

> “witch hunt” is most often used by bad actors attempting to discredit legitimate investigations).

By that logic, "fake news" is now unusable because Trump weaponized it, despite the term accurately describing a real phenomenon that existed before and after his usage. "Gaslighting" would be suspect because it got picked up by people dramatizing ordinary disagreements. Every useful term for describing social dynamics gets captured by someone with an agenda eventually.

Hitler liked chocolate, doesn't mean you shouldn't eat chocolate. "You used a word that bad people also use" is not interesting - it's a way of avoiding the object-level debate while still claiming moral high ground.


> "Gaslighting" would be suspect

Gaslighting would be simply incorrect, since gaslighting refers to an elaborate scheme of making somebody doubt their own perception/sanity. It a a severe form of abuse, requires an ongoing relationship with power dynamics (it cannot happen from a single instance of interaction), and typically results in long-term PTSD for the victim(s).

Agree on the capturing. Watering down terms is highly unfortunate for everyone.


> Every useful term for describing social dynamics gets captured by someone with an agenda eventually.

So, in essence, you’re agreeing.

> Hitler liked chocolate, doesn't mean you shouldn't eat chocolate.

Arguments have nothing to do with dietary preferences, that comparison makes no sense.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: