Hacker News new | ask | show | jobs
by JeremyNT 1054 days ago
The fact that Meta is able to lie and call their restrictive licensing open source is nearly as misleading as "OpenAI."

We need to do better than to repeat these claims uncritically. The weight licenses are not "open source" by any useful definition, and we should not give Meta kudos for their misleading PR (especially considering that they almost surely ignored any copyright when training these things - rules for thee, but not for me).

"Not as closed as OpenAI" is accurate, but also damning with faint praise.

2 comments

Just some general piece of advice: it's not productive to constantly be giving out the worst criticism you possibly can when someone does something that's not terrible but still unacceptable. Doing so just tells the companies that nothing satisfies the community and that they should stop trying. Instead, it's better to mention what they did right and point to how they can make it better.
Can you chill? It’s def open source
The source code is, as it's MIT, but the weights are not, as they're CC-BY-NC: https://github.com/facebookresearch/audiocraft#license

  about: pytorch @ fb.
So I can build a business on it, then?
I believe Meta has explicitly said that you can, but that's not what open source means and the model isn't open source.
Meta says to imagine you can: "Imagine a professional musician being able to explore new compositions without having to play a single note on an instrument. Or an indie game developer populating virtual worlds with realistic sound effects and ambient noise on a shoestring budget. Or a small business owner adding a soundtrack to their latest Instagram post with ease."

In reality, you can't, as they licensed the weights for noncommercial use only: https://github.com/facebookresearch/audiocraft#license

Research does exist you know. This is immensely helpful for a huge number of people in academia.

If you want to build a company, perhaps you should do what everyone in the industry has done for millennia, copy the movements performed and optimize them while doing so.