I don't think it's freedom tech if we depend on these companies that own the large data sets to train the base model. Also, the only reason llama is 'free' right now is someone just started handing it out when they weren't supposed to, the license wasn't really that permissive. And it doesn't matter anyway because you wouldn't be able to train your own unless you have big tech access to big data so that means those companies will always have the upper hand and can hold back newer versions at will. Same trap as before, 'free'.. for now..

But ya, it's still pretty cool tech. Start hoarding that data now or it'll be just like Google search and their web crawler monopoly.