Most answers in this thread are either completely misinformed or based on unfounded speculation.
If we want to avoid guessing at “evil mastermind” plots that might or might not be there, the answer can actually be a pretty simple one: same reason why they (and similarly large companies like Microsoft and Google) contribute to so many open source projects, and publish a big chunk of their research in scientific journals and conferences. Collaboration and sharing development of certain technologies can be good and profitable in the long run.
Especially when it comes to AI research, no single institution in the world is able to “do it alone”, or to get significantly ahead of the others, because of how hard it is, how expensive it is, and how few people have the necessary know-how and ingenuity to effectively contribute. Sharing LLAMA is then probably intended as a way to boost research by third parties, especially universities and the like, helping new people / ideas get into the game, which can then be recruited / adopted by Meta for their own products.
Latest Answers