ParserComp 2025 - Initial Results

[pieartsy (Aster F)]

@ChristopherMerriner

The game authors are not at fault for this.

I am inclined to disagree with this assessment. I don’t at all blame you or fos1 for making your decision to accept the games based on the information the authors provided to you:

@ChristopherMerriner

the code and narrative of the game would be human-created, while the parser mechanism relied on AI

@mmarubio

given the extensive human-written basis that is required to make something like this work - they felt we fit within the scope.

But that information turned out to be demonstrably and flagrantly false. It’s in the games themselves (see pinkunz’s screenshots)-- they say they were made with AI as a “parser/translator” but are pinging a server live and generating gobs of text-book AI writing. That’s not your fault, that’s mischaracterization on the authors’ part.

I do think that, upon realizing this misrepresentation, the games should have been disqualified afterward without fears of reneging on a previous decision or somesuch, especially since it was made on (unbeknownst to you) false premises. But the original decision is not something to feel bad about, imo.