3 ms·
For being an article about "open source" and the implications on geo-politics, it seems like a not sufficiently researched article. Something obvious that stick
by diggan 2y ago
For being an article about "open source" and the implications on geo-politics, it seems like a not sufficiently researched article. Something obvious that sticks out:
> Meta & Llama: In 2023, Meta was behind OpenAI and Google in AI. By releasing Llama openly, Meta gained community support and adoption, striking at rivals’ closed approaches.
Meta might call Llama "open source" in their marketing material, but all their legal documents call Llama "proprietary" and anyone taking a closer look at their license/use policy would immediately be able to point out how Llama isn't open source.
Since the article gets something really basic wrong, how could we trust the rest of the article?
- giuliomagnifico 2y agoThe article talks about the fact that those model are "available to anyone", even Llama is Open Source from that point of view: https://github.com/meta-llama/llama-models https://github.com/meta-llama/llama-models Where Llama is not Open Source is in total respect of certain ethical principles: https://opensource.org/blog/metas-llama-license-is-still-not-open-source https://opensource.org/blog/metas-llama-license-is-still-not... So, it’snot 100% Open Source inside the context of the article it’s acceptable. Moreover, even DeepSeek is not OpenSource, it’s "OpenWeight"
- jraph 2y agoWords have meaning and sometimes it's important to remain clear. I think it is not acceptable to use "open source" for things that are not actually "open source", especially in an article about the topic. There are probably better words to describe what needs to be described instead. See my other comment [1] where I try to defend why I think like this. (Now, as I'm saying in the other comment, in this case for this article, it's just an example that doesn't work, it's not something fundamentally wrong about it.) [1] https://news.ycombinator.com/item?id=43613453#43614394 https://news.ycombinator.com/item?id=43613453#43614394
- 9283409232 2y agoThis is the type of thing that only people who get in internet arguments about MIT vs BSD care about. Is the llama license open source as defined by the OSD? No. But for the majority of people, they don't care. For all intents and purposes, it is open source for them. All of this is to say that I don't fault the author for just calling it open source.
- selfhoster11 2y agoExcept that Llama 4's license restricts you from using its multimodal features in the European union, and the Llama 3 license restricted using the model output to augment the training of other models (distillation). Both are important use cases. It very much isn't an academic exercise. It has a real impact on what you can (legally) do with a model. With a model licensed using MIT, you know what you are getting. You know you can modify it. You know you can use it for any purpose whatsoever, including ones the corporation that built it dislikes. Using the MIT license is "no ace up my sleeve" licensing - there are no gotchas that will come and bite you later.
- diggan 2y ago> Except that Llama 4's license restricts you from using its multimodal features in the European union The EU restrictions were added in 3.2's Use Policy, while 3.1 License introduced the restrictions on using model output for training other models. I've collated the changes here: https://notes.victor.earth/how-llamas-licenses-have-evolved-over-time/ https://notes.victor.earth/how-llamas-licenses-have-evolved-... Besides that, I agree with what you're saying.
- selfhoster11 2y agoI appreciate the correction. I've not been paying much attention to Llama licensing after the first Llama 3 models came out.
- pabs3 2y agoModel output isn't copyrightable, and I would be surprised if them models themselves were copyrightable too, which means the licenses are meaningless.