3 ms·
This is a really interesting move — as others have been saying in the comments I wonder how AI startups will react going forward. I bet there will be a lot mor
by lucasfcosta 3y ago
This is a really interesting move — as others have been saying in the comments I wonder how AI startups will react going forward.
I bet there will be a lot more effort to build truly open-source models. Also, I wonder why no foundation has yet got involved to pool resources and create large enough models.
It feels like AI needs its own "CNCF".
- flakiness 3y agoMeta is inching towards to that direction through llama series as Google did it through Kubernetes. It'll happen once the dusts settle. I'm very impressed by how Meta position itself as "AI for rest of us" position through llama and to an extent PyTorch, although I have no idea how they are going to capitalize that position besides hiring. (vs. Google having a cloud offering.)
- lucasfcosta 3y agoThe problem with llama is that it's not really open. So it's not really "for the rest of us". It's just "for the rest of us" until we reach a significant number of users and then Meta is gonna come for us.
- dannyw 3y agoIt's 700 million monthly active users on the LLaMA 2 release date; so if you reach and exceed a significant number of users in the future, you don't lose your license. As a purist do I wish it was fully open source? Yes. Is it restrictive to "the rest of us"? No. Finally, if you have more than 700M MAUs, you probably have an internal LLM you should be using.