From f3161c28b09b77b04eb3a0d077603066cc52433a Mon Sep 17 00:00:00 2001 From: HJfod <60038575+HJfod@users.noreply.github.com> Date: Fri, 19 Apr 2024 11:47:41 +0300 Subject: [PATCH] add a TL;DR to AI guidelines --- mods/guidelines.md | 2 ++ 1 file changed, 2 insertions(+) diff --git a/mods/guidelines.md b/mods/guidelines.md index ea9ef38..cbeee1d 100644 --- a/mods/guidelines.md +++ b/mods/guidelines.md @@ -49,6 +49,8 @@ These are some things one might expect to result in a rejection, but actually mo Due to controversy among developers regarding the subject, mods that utilize **Generative AI** have to follow these additional guidelines in relation to their AI usage. +> :information_source: TL;DR; All mods using AI must be trained on consentually obtained data, and may not be used for creating levels. + * The mod **may not grant an unfair advantage** in gameplay or anything else to people through the use of AI * The mod **may not compromise the integrity of user-created content**. In essence, this means that **no mod is allowed to use AI for placing, modifying, or deleting objects**, though it is not limited to just those. * The AI model is only, and as provably as one can reasonably expect, **trained on consentually and transparently obtained data**. Any mod found to be breaking this rule would be found to be in violation of the "collecting user data without consent" hard-ban rule, and subsequently result in the developer being immediately and permanently banned from the index