Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Open-sourcing AudioCraft: Generative AI for audio

ai.meta.com

Post reply on HN
Print view View original

81–90 of 335 posts

Re: Open-sourcing AudioCraft: Generative AI for audio

#81

Post by kmeisthax »

eropple wrote: post #34
kmeisthax wrote: post #27

Earlier quoted context omitted.

This is purely a function of everyone remembering the RIAA's decade-long campaign to prevent people from taking the music they had rightfully stolen. As far as I'm aware LLaMA was trained on "publicly available data"[0], not "licensed data". Furthermore, MusicGen's weights are licensed CC-BY-NC, which is effectively a nonlicense as there is no noncommercial use you could make of an art generator[1]. This is not only…

> MusicGen's weights are licensed CC-BY-NC, which is effectively a nonlicense as there is no noncommercial use you could make of an art generator How do you figure? Have you never just...made stuff to make stuff?

In copyright law the use of the work itself is considered a commercial benefit, so "noncommercial use" is an oxymoron. Consider these situations:

- If I use AudioCraft to post freely-downloadable tracks on my SoundCloud, I still get the benefit of having a large audio catalog in my name, even if I'm not selling the individual tracks. I could later compose tracks on my own and ride off the exposure I got from posting "noncommercially".

- If I run AudioCraft as a background music generator in my store, I save money by not having to license music for public performance.

- If I host AudioCraft on a website and put ads on it, I'm making money by making the work available, even though I'm not charging a fee for entry.

I suspect that a lot of people reading this are going to have different arguments for each. My point is that if you don't think that all of these situations are equally infringing of CC-BY-NC, then you need to explain why some are commercial and some are not. Keep in mind that every exception you make can be easily exploited to strip the NC clause off of the license.

If you're angry at the logic on display here, keep in mind that this is how judges will construe the license, and probably also how Facebook will if you find a way to make any use of their AI. The only thing that stops them from rugpulling you later is explicit guidance in CC-BY-NC. Unfortunately, the only such guidance is that they don't consider P2P filesharing to be a commercial use.

So, absent any other clarifications from Facebook, all you can do without risking a lawsuit is share the weights on BitTorrent.

EDIT: And yes, I have made stuff just to make stuff. I license all of that under copyleft licenses because they express the underlying idea of 'noncommercial' better than actual noncommercial clauses do.

Re: Open-sourcing AudioCraft: Generative AI for audio

#82

Post by cooper_ganglia »

gilmore606 wrote: post #72
Pannoniae wrote: post #69

"generating new music in the style of existing music" will probably be a huge field soon. I can't wait for it to happen, it's a low-cost way of producing even more music to listen to.

> I can't wait for it to happen, it's a low-cost way of producing even more music to listen to. I can't really understand this. I'm a DJ and a huge music nerd, and I spend a lot of time every week discovering new music from the past 100 years and all over the world, and I'm constantly struck by _how much of it there is_. I've spent weeks just digging through psych-funk records from West Africa from the 1970s. How can…

Music is self-expression. I don’t always identify entirely with others. I always identify with my self. Having music generated for you on such a personalized level is an attractive prospect.

I don’t think this replaces “100% organic, human-made” music, though. I think there’ll always be a reason to listen to music made by other people. But I think this changes the landscape of how and why people create music to begin with. It certainly will devalue existing music, since everyone has something they may prefer that they can generate instantly.

I think generative AI is a terrible technology for artists who want to make money from their art, but in my personal opinion, I strive for a world where art isn’t a transaction, but a gift of human expression and connection. A world where art is appreciated for the emotion, stories, and ideas it conveys rather than the monetary value it holds. Generative AI might disrupt the traditional economic models in the art world, but it also opens up new opportunities for creative exploration and personal expression. It’s a challenging evolution, but one that could potentially democratize art, making it more accessible and personal than ever before! Bring on the Renaissance: Part 2!

Re: Open-sourcing AudioCraft: Generative AI for audio

#83

Post by skybrian »

As an amateur musician I’m wondering if there are any of these audio generators that you can give a tune or chord progression to riff on. ABC format maybe? There are lots of folk tunes on thesession.org.

Could you generate a rhythm track? Ideally you could make songs one track at a time, by giving it a mix of the previous tracks and asking it to make another track for an instrument. Or, give it a track and ask it to do some kind of effect on it.

Another interesting use might be generating sound samples for a sampled instrument.

Re: Open-sourcing AudioCraft: Generative AI for audio

#84

Post by archontes »

gavman wrote: post #13

> MusicGen, which was trained with Meta-owned and specifically licensed music, generates music from text-based user inputs, while AudioGen, which was trained on public sound effects, generates audio from text-based user inputs. Meta is really clearly trying to differentiate themselves from OpenAI here. Open source + driving home "we don't use data we haven't paid for / don't own".

You don't own data. You can sometimes copyright data.

https://www.americanbar.org/groups/science_technology/public...

Re: Open-sourcing AudioCraft: Generative AI for audio

#86

Post by rideontime »

cooper_ganglia wrote: post #82
gilmore606 wrote: post #72

Earlier quoted context omitted.

> I can't wait for it to happen, it's a low-cost way of producing even more music to listen to. I can't really understand this. I'm a DJ and a huge music nerd, and I spend a lot of time every week discovering new music from the past 100 years and all over the world, and I'm constantly struck by _how much of it there is_. I've spent weeks just digging through psych-funk records from West Africa from the 1970s. How can…

Music is self-expression. I don’t always identify entirely with others. I always identify with my self. Having music generated for you on such a personalized level is an attractive prospect. I don’t think this replaces “100% organic, human-made” music, though. I think there’ll always be a reason to listen to music made by other people. But I think this changes the landscape of how and why people create music to begin…

> I strive for a world where art isn’t a transaction, but a gift of human expression and connection. A world where art is appreciated for the emotion, stories, and ideas it conveys rather than the monetary value it holds.

In a world where nobody is compensated for their art, the only people making art will be the ones privileged enough to have the means to do so for free. I don't see how this leads to "Renaissance: Part 2."

Re: Open-sourcing AudioCraft: Generative AI for audio

#89

Post by smoldesu »

ChildOfChaos wrote: post #58
gavman wrote: post #13

> MusicGen, which was trained with Meta-owned and specifically licensed music, generates music from text-based user inputs, while AudioGen, which was trained on public sound effects, generates audio from text-based user inputs. Meta is really clearly trying to differentiate themselves from OpenAI here. Open source + driving home "we don't use data we haven't paid for / don't own".

It's likely partly a PR/branding exercise as well. In the new world that Meta sees, of VR/AR and AI, Meta is in a position already were people don't want them to have much power in this world, because they don't trust them over privacy etc, meta is trying to pivot to become more trustworthy so they make genuine moves in this space.

That, or this is an ongoing research lab (FAIR) that has existed for ~half a decade and has advanced the state-of-the-art in AI further than Apple, Microsoft and Google combined.

Re: Open-sourcing AudioCraft: Generative AI for audio

#90

Post by CharlesW »

smallerfish wrote: post #78

This is Spotify's route to profitability - the Netflix model of generating their own "content" (/music), and not having to pay the labels. Premium plans for us music nerds who want a human at the other end, regular plans for plebs who just want to fill the silence with something agreeable.

Although I think AI-generated and AI-augmented (using voice cloning, etc.) artists are a given, for Spotify to stop paying labels they'd have to be able to remove all non-Spotify content from their streaming catalog. That doesn't seem like a possibility in our lifetimes. (Also, Spotify hasn't even been able to build a sustainable business on podcasts, which they copy to their closed platform for free.)

It's an interesting thought experiment, though. I can imagine that "environmental audio" companies like Muzak have about 5 years left before they either adapt or die. What other kinds of companies are in trouble?

Post reply on HN