Conversation

Notices

1. Embed this notice
   翠星石 (suiseiseki@freesoftwareextremist.com)'s status on Wednesday, 28-May-2025 00:12:56 JST 翠星石
   @halva All LLM's are inherently proprietary, as no source code results from training - only object code.
   
   Furthermore, all current models constitute copyright infringement, due to removal of copyright information and mixing of incompatible licenses (although it would be possible to train on texts that are under compatible licenses).
   
   It clearly not possible to validly license a model, but so far only a handful have been released under a free license.
   • Embed this notice
     翠星石 (suiseiseki@freesoftwareextremist.com)'s status on Friday, 06-Jun-2025 20:41:28 JST 翠星石

     in reply to

     • 👺防空識別區👹
     • マルピエロ
     @mavalalkujuuu @adiz The concept of imaginary property is indeed proprietary; https://www.gnu.org/philosophy/not-ipr.html
     
     Copyright by default makes things proprietary, but you can use it against itself to make things always free.
   • Embed this notice
     マルピエロ (mavalalkujuuu@fsebugoutzone.org)'s status on Friday, 06-Jun-2025 20:41:29 JST マルピエロ

     in reply to

     • 👺防空識別區👹
     @adiz @Suiseiseki intellectual property and copyright are proprietary.
     SuperDicq repeated this.
   • Embed this notice
     👺防空識別區👹 (adiz@mtl.jinxian.casa)'s status on Friday, 06-Jun-2025 20:41:30 JST 👺防空識別區👹

     in reply to

     @Suiseiseki intellectual property doesn't exist and copyright laws are dumb. @halva