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

281–290 of 335 posts

Re: Open-sourcing AudioCraft: Generative AI for audio

#281

Post by d0odk »

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…

if you want to express yourself musically, then learn how to play an instrument and compose music

Re: Open-sourcing AudioCraft: Generative AI for audio

#282

Post by waffletower »

I think it is a mistake to acquiesce and let copyright owners bully AI model trainers over model data inputs. The endgame of this practice is a "pay per thought" society. This is separate from speculation regarding machine sentience -- as interfaces improve AI models will serve more and more as direct human extensions of mind. While copyright duration is a separate issue, and the current durations are appalling, copyright violations should focus strictly upon the output of models and how they are utilized. There are so many melodies in my head that I have not nor will I ever pay for, (some of which I would love to remove). AI models need also to have the same unfettered access to the commons as we do. Infringement occurs on the outputs -- application of copyright restrictions on model inputs is a violation of Fair Use and a definite money grab.

Re: Open-sourcing AudioCraft: Generative AI for audio

#284

Post by painted-now »

trojan13 wrote: post #21

Finally, a way to fulfill my childhood dream of composing a symphony of rubber ducks honking. Bach would be proud. /edit On a more serious node. I already see the 24/7 lofi girl streaming generated music. The sample[1] on lofi sounds pretty good. [1] https://dl.fbaipublicfiles.com/audiocraft/webpage/public/ass... "Lofi slow bpm electro chill with organic samples"

I also like some of the generated examples.

Can I haz full version of Bach + `An energetic hip-hop music piece, with synth sounds and strong bass. There is a rhythmic hi-hat patten in the drums.` please?

(https://dl.fbaipublicfiles.com/audiocraft/webpage/public/ass...) ?

Re: Open-sourcing AudioCraft: Generative AI for audio

#285

Post by chrisjj »

emporas wrote: post #272
chrisjj wrote: post #218

Earlier quoted context omitted.

Evidence? None of the demos suggest that is true.

Did they change the base model? If not, then the audiocraft_plus, is based on audiocraft, and it creates music of length close to 5 minutes. I don't know if audiocraft_plus incorporates all three modalities of the release, MusicGen, AudioGen, and EnCodec. It uses MusicGen for sure, all four models, small, medium, large and melody. https://github.com/GrandaddyShmax/audiocraft_plus

Bit is that "extension" to 5mkns more than just the repeat of e.g. 15secs heard in the demos?

Re: Open-sourcing AudioCraft: Generative AI for audio

#286

Post by makestuff »

These models are going to end up being used for advertising. Soon pretty much every ad you see will be generative AI based. It makes A/B testing way easier as you no longer need a creative person to modify the ad or change something subtle about it. For example, the generative voice might change to a different speaker or something, and the AI can generate thousands of different voices to see which one is most effective.

Re: Open-sourcing AudioCraft: Generative AI for audio

#287

Post by emporas »

chrisjj wrote: post #285
emporas wrote: post #272

Earlier quoted context omitted.

Did they change the base model? If not, then the audiocraft_plus, is based on audiocraft, and it creates music of length close to 5 minutes. I don't know if audiocraft_plus incorporates all three modalities of the release, MusicGen, AudioGen, and EnCodec. It uses MusicGen for sure, all four models, small, medium, large and melody. https://github.com/GrandaddyShmax/audiocraft_plus

Bit is that "extension" to 5mkns more than just the repeat of e.g. 15secs heard in the demos?

Yes, of course it is more than that. You can create full songs of several minutes. I have published 20 songs on youtube of 2-3 minutes each. Listen to that song, the best i have made so far [1]. It doesn't repeat at all!

I haven't looked closely to this release, is the audiocraft page on github different than facebookresearch/audiocraft? The other two modalities may be new, AudioGen, and EnCodec, but i was under the impression that they changed the license to full open source, and that was that.

https://www.youtube.com/watch?v=UL1KmrHMjcM

Re: Open-sourcing AudioCraft: Generative AI for audio

#288

Post by tiahura »

kmeisthax wrote: post #187
stale2002 wrote: post #106

Earlier quoted context omitted.

This is a weird comment. Do you think that non commercial use simply doesn't exist or something? Because non commercial use isn't some crazy concept. It is a well established one, that doesnt disclude literally everything. Also, you are ignoring the idea that Facebook will almost certainly not sue anyone for using this for any reason, except possibly Google or Apple. So if you aren't literally one of those companies…

Noncommercial use is not well established in copyright law, which is the law that actually matters. I know other forms of law actually do establish noncommercial and commercial use standards, but copyright does not recognize them. As for "Facebook won't sue"? Sure, except we don't have to worry about just Facebook. We have to worry about anyone with a derivative model. There's an entire industry of copyleft trolls[0]…

Commercial vs Noncommercial use is well established in copyright law - in everything from Final Rule Regarding the Noncommercial Use Exception to Unauthorized Uses of Pre-1972 Sound Recordings https://www.copyright.gov/rulemaking/pre1972-soundrecordings... to Noncommercial webcasters https://www.law.cornell.edu/uscode/text/17/114#f_4 to Fair Use.

Noncommercial licenses are taken up in "GREAT MINDS v. FEDEX OFFICE AND PRINT SERVICES, INC 886 F.3d 91 (2nd Cir. 2018). Thé court explains they are enforceable and are basically just a category of contract. So, as long as the contract is clear, it’s probably enforceable.

Re: Open-sourcing AudioCraft: Generative AI for audio

#289

Post by smoldesu »

lannisterstark wrote: post #221
RyanAdamas wrote: post #152

For my two cents; the goal of our human pursuit is to advance our technologies and systems to the point we can all live carefree lives where the focus of our pursuits is self defined. It is not to "own more rights" to shit. Sorry, but no one actually owns what they create, that's the point of creation. And if you take issue with the term create, that only reinforces my point. We're all influence machines, input and o…

Ah yes, you should be free to rewrite the fictional work I wrote or add one chapter to it and be free to sell it under your name magically implying that you are the author. Screw the original artists, right? Why should they deserve anything. Thankfully your opinion is an extreme opinion and will never come to pass. Tough shit indeed. :) ---- I really like hackernews but recently I've been seeing a plethora of "your r…

I'm able to do everything in your hypothetical scenario besides illegally copy and redistribute the original content. If it makes me happy to write my name on the cover and pretend I wrote "Robert Frost's Poetry Collection" then so be it. Truly, if I want to say "fuck the original artists" in the comfort of my home or margins of the pages, I can do so. I can even sell that adulterated copy under the First Sale doctrine.

> I've been seeing a plethora of "your rights don't matter, you own nothing" bs spreading around.

It's not a new sentiment. Plenty of countries worldwide ignore US copyright, if they all got sanctioned then America wouldn't have have electricity.

If you author a copy of your content in a digital medium, you should be prepared for that content to be redistributed against your will, infinitely. It's not the nature of humanity, it's the nature of the digital format.

Post reply on HN