- cross-posted to:
- hackernews
- cross-posted to:
- hackernews
You must log in or register to comment.
Just remember to try really hard to not to seed it and say it’s training data… And it’s fair use.
Wasn’t it just meta data?
So far.
These millions of audio files have done nothing wrong. Keeping them locked away is scandalous. Release them immediately !
/dad joke, sorryHow dare they, someone should call, PETM. People for the ethical treatment of music. Those poor songs locked up all day.
I’m sure it’s been scraped plenty of times by AI companies who are doing way more damage.
Yeah but that’s damage to artists. AI music gives Spotify something to put into a playlist that they don’t have to pay even their meagre rate to.
It hurts record companies. They want to own all AI generated music. It’s quite clear with what happened to udio. It’s monopolies against open source, not AI against artists.
I used some software to download music from them and they locked me out of my account for violating their ToS.
Since day one, we have stood with the artist community against piracy, and we are actively working with our industry partners to protect creators and defend their rights.
Lmao. They protect and defend artist rights so hard they they’ve refused to pay a fair compensation, and have taken it further by promoting AI artists over actual artists. This statement is almost comical after it was reported that they’ve had copy cats to replace King Glizard and the Lizard Wizard when they pulled their albums from Spotify
It have never been a fight against privacy, it’s a fight against freedom
Any sauce on the KGLW story? Had not heard of this.
Thank you!
Spotify absolutely deserves to be singled out for its exploitative practices, especially since this company is largely responsible for musicians not being paid fairly for their hard work. It’s just a shame that there’s hardly anything to steal here other than people’s hard work, to which Spotify has contributed nothing - but that applies to all companies that are successful on the internet today. Without exception, all of these companies are built on the same platform logic: the content that these companies exploit is paid for with starvation wages, if at all (not at all in the case of LLMs).
Therefore, I cannot see anything positive in this because it does not change the underlying problem in the slightest.
The major labels are still the biggest evil
Spotify & Co. make advance payments to the labels to be allowed to use their music catalogues. These advance payments are then recouped with the streaming revenues. However, if the revenue is less than the advance, the difference remains with the labels as “breakage”. If a streaming service pays a label US $1 million as an advance for the contract period, but the label’s catalogue is only streamed to the value of US $750,000, then the label has US $250,000 in additional revenue that does not have to be distributed to the artists.
Nevertheless, Spotify makes more profit than any music label, even more than all the remaining music labels combined. This is how it works today: music, literature, journalism, and art no longer exist according to this logic - only content. And as disrespectful as the term sounds, that’s how it’s paid for - with scrabs because that’s the business model.
Your pirate approach is no longer up to date, because it is no longer directed against large corporations, but robs artists of the little they have left. This will only accelerate the trend: no one will try to make a living from art anymore. If you think that people will do it anyway because they want to express themselves, I think you are absolutely wrong.
It’s not accelerating the trend one bit by opening it to everyone. Music labels and Spotify don’t plan on putting a stop to AI, they want to own it. The artists lost decades ago and siding with copyright juggernauts doesn’t help anyone but the copyright juggernauts.
It makes more profits but not revenus
Market cap and profits are different metrics. At the end of the day major labels dictate streaming services policies . Considering that they own 80% of all music
I thought it was gonna be some no-name group that was gonna hold it ransom etc. But it’s actually by Anna’s-archive. I don’t really condone piracy (pay the people who make art, and those who make it accessible), but if anyone was going to do it, I’m glad it’s them.
I don’t support piracy either. I just torrent it for free. 🏴☠️
Lars Ulrich is raging,he’s gonna sue.
gonna have to wait an extra 2 months to get that gold plated shark tank.
The sharks are gonna be so bummed, I know they had their hearts set on it.
Where will all of this be posted? … So that I can avoid it, ofc.
Anna’s
“… Found to be 48% AI Slop”
I hope musicbrainz hops on the metadata list
Eww, why scrape that cesspool of AI generated slop?
One thing I’ve seen people talking about is the metadata library, which is apparently very good
Does it have vector embeddings of all the songs if they did it would be great for building a foss recommendation algorithm
Because not all of Spotify is AI slop.
Not all. Just a huge chunk.
The amount of “AI slop” is an incredibly tiny portion of the metadata.
Because before the AI slop there was real content that was only ever released to Spotify.
