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

131–140 of 335 posts

Re: Open-sourcing AudioCraft: Generative AI for audio

#131

Post by crazygringo »

ysleepy wrote: post #126
crazygringo wrote: post #114

Earlier quoted context omitted.

No. We've had photo and audio manipulation for many decades now. For a long time now, we've had to separate out what's credible from what's bullshit. Fortunately, it's pretty simple in real life. We have certain publications and sources we trust, whether they're the NYT or a respected industry blog. We know they take accurate reporting seriously, fire journalists who are caught fabricating things, etc. If we see a cl…

I don't agree. With ML tools it is possible to make sweeping changes to images and text that are often impossible to detect. combined with the centralisation of most online activities, large players could alter the past. Imagine facebook decides to subtly change every public post and comment to show some particular person or cause in a better light.

If one "large player" like the NYT decides to "alter the past", you can compare with the WaPo or any other newspaper. You can compare with the Internet Archive. You can compare with microfiche. These aren't "impossible to detect", they're trivial to detect if you bother to compare.

We have tons of credible archived sources owned by different institutions. And these sources are successful in large part due to their credibility and trustworthiness.

It's just not economically rational for any of them to start "altering the past", and if they did, they'd be caught basically immediately and their reputation would be ruined.

This isn't an ML/tooling question, it's a question of humans and reputation and economic incentives.

Re: Open-sourcing AudioCraft: Generative AI for audio

#132

Post by amelius »

crazygringo wrote: post #114
bottlepalm wrote: post #102

Anyone feel like with the flood of AI generated content there's a risk of the past being 'erased'. Like in 10 years we won't be able to tell if any information from the past is real or fake - sounds, pictures, videos, etc.. Like we need to start cryptographically signing all content now if there's any hope of being able to verify it as 'real' 10 years from now.

No. We've had photo and audio manipulation for many decades now. For a long time now, we've had to separate out what's credible from what's bullshit. Fortunately, it's pretty simple in real life. We have certain publications and sources we trust, whether they're the NYT or a respected industry blog. We know they take accurate reporting seriously, fire journalists who are caught fabricating things, etc. If we see a cl…

> We've had photo and audio manipulation for many decades now. For a long time now, we've had to separate out what's credible from what's bullshit.

The difference is that the floodgates are being opened.

Re: Open-sourcing AudioCraft: Generative AI for audio

#133

Post by TrackerFF »

jeffbee wrote: post #116

The fact that it generates a song for the prompt "Earthy tones, environmentally conscious ... organic instrumentation" goes a long way to proving that English words no longer mean anything particularly.

I've played guitar 25 years, and it's funny how the music community has been using all kind of words to describe music or tone. Describing certain tone as "hairy", "mushy", "wooly", "airy", "buttery", etc. is just very common.

Re: Open-sourcing AudioCraft: Generative AI for audio

#134

Post by hereonout2 »

s1k3s wrote: post #4

The demos are great. Could someone explain what’s in it for Meta open sourcing all these models?

Was asking myself the same earlier, I'm sure it is largely to do with publicity and the fact that selling these services is not their core business. At the very least releasing this stuff probably won't damage their core business but will take the sheen off of some other big names.

I wondered though, generative AI is hurling us into a world where we'll need more mechanisms to sort real from fake, provenance will play a large part, and meta's platforms could be part of the answer. i.e. content linked to actual verifiable people.

Re: Open-sourcing AudioCraft: Generative AI for audio

#136

Post by russdill »

bottlepalm wrote: post #102

Anyone feel like with the flood of AI generated content there's a risk of the past being 'erased'. Like in 10 years we won't be able to tell if any information from the past is real or fake - sounds, pictures, videos, etc.. Like we need to start cryptographically signing all content now if there's any hope of being able to verify it as 'real' 10 years from now.

If you're watching a movie or TV show, a vast majority of the sounds you are hearing are not "real". Has that bothered you before?

Re: Open-sourcing AudioCraft: Generative AI for audio

#139

Post by probablynish »

crazygringo wrote: post #114
bottlepalm wrote: post #102

Anyone feel like with the flood of AI generated content there's a risk of the past being 'erased'. Like in 10 years we won't be able to tell if any information from the past is real or fake - sounds, pictures, videos, etc.. Like we need to start cryptographically signing all content now if there's any hope of being able to verify it as 'real' 10 years from now.

No. We've had photo and audio manipulation for many decades now. For a long time now, we've had to separate out what's credible from what's bullshit. Fortunately, it's pretty simple in real life. We have certain publications and sources we trust, whether they're the NYT or a respected industry blog. We know they take accurate reporting seriously, fire journalists who are caught fabricating things, etc. If we see a cl…

Seems like it might now become much easier to post a clip on YouTube that looks like an authentic BBC clip, logo and all. If generative AI gets that good, how will you be able to tell whether a particular piece of media comes from a trusted source?

Might not be possible on platforms - only if it's posted on a trusted domain.

Re: Open-sourcing AudioCraft: Generative AI for audio

#140

Post by conductr »

amelius wrote: post #132
crazygringo wrote: post #114

Earlier quoted context omitted.

No. We've had photo and audio manipulation for many decades now. For a long time now, we've had to separate out what's credible from what's bullshit. Fortunately, it's pretty simple in real life. We have certain publications and sources we trust, whether they're the NYT or a respected industry blog. We know they take accurate reporting seriously, fire journalists who are caught fabricating things, etc. If we see a cl…

> We've had photo and audio manipulation for many decades now. For a long time now, we've had to separate out what's credible from what's bullshit. The difference is that the floodgates are being opened.

At a time when “people” seem easily manipulated and focused on their fully believing their personal feeds of curated outrage. They often don’t apply the screens/filters they should be because of the apparent social proofs, trust, and biases they have with the content. Contemporary journalists hardly do any fact/source checks as it is. So they’ll begin reporting on some of this, giving it further credibility and it’s just a downward spiral. So, more of the same, yay!
Post reply on HN